Embed Size (px)
Citation preview
INTERNATIONAL MONETARY FUND
MEXICO
Selected Issues
Prepared by Martin Kaufman, Esteban Vesperoni, Herman Kamil, Jeremy Zook (all WHD), Santiago Acosta, Enrique Flores (both FAD), Pamela Madrid (MCM), Gilda
Fernandez (SPR), and Francisco Vazquez-Ahued
Approved by Western Hemisphere Department
November 1, 2012 Contents Page
I. Mexico: A Closer Look at Global Spillover Channels ............................................................................... 4 A. Introduction ................................................................................................................................. 4 B. Mexico’s Linkages with the U.S. and the Rest of the World ....................................................... 4 C. Sensitivity of Mexico’s Growth and Asset Prices to Global Shocks ......................................... 11 D. Conclusion ................................................................................................................................. 14 References ..................................................................................................................................... 15
II. What Explains Mexico’s Recovery of U.S. Market Share? ................................................................... 16
A. Background and Recent Developments .................................................................................. 16 B. Evolution of U.S. Market Share: Mexico and China .............................................................. 19 C. What Explains Mexico’s Increased Competitiveness? ........................................................... 21 References ..................................................................................................................................... 25
III. The Case for Tax Revenue Mobilization in Mexico ............................................................................. 29
A. Introduction ............................................................................................................................ 29 B. Macroeconomic Considerations for Revenue Mobilization ................................................... 29 C. Mexico’s Tax System: A Cross-Country Perspective ............................................................ 31 D. Overview of Mexico’s Tax System ......................................................................................... 34 E. Options to Strengthen Revenue Mobilization ......................................................................... 40 References .....................................................................................................................................43
IV. Mexico: Financial Sector Development ................................................................................................ 45
A. Background .............................................................................................................................. 45 B. Financial Development: Mexico in the Context of Emerging Markets .................................. 48 References ..................................................................................................................................... 49
2
V. Mexico: Migration and Labor Markets .................................................................................................. 57 A. Background ............................................................................................................................... 57 B. Recent Developments in Labor Markets ................................................................................... 58 C. Migration: Recent Developments ............................................................................................. 59
D. Migration and Labor Markets .................................................................................................... 61 References ..................................................................................................................................... 62
Tables I.1. Top Mexico Export Destinations ..................................................................................................... 5 I.2. Mexico and US GDP Cointegration Post-NAFTA .......................................................................... 5 I.3. Matrix of Correlation Coefficients .................................................................................................. 6 I.4. Variance Decompositions from Bivariate VARs: Mexico Real GDP ............................................. 6 I.5. Variance Decompositions from Bivariate VARs: Mexico Real Exports ......................................... 6 I.6. Real Export Growth by Region ........................................................................................................ 7 I.7. Announced Plans for Future FDI in Mexico, 2012 .......................................................................... 7 I.8. Top Ten Sources of FDI, 2010 ......................................................................................................... 9 I.9. Top Ten Destinations of FDI, 2010 .................................................................................................. 9 I.10. Stock of Portfolio Liabilities by Region ........................................................................................... 9 I.11. Stock of Domestic Public Sector Debt Held by Foreigners ........................................................... 11 II.1. Sectors with the Biggest gain in U.S. Import Market Share: 2005–2012 ...................................... 18 II.2. Sectors in Which the Largest Fraction of Market Share Gain Accrued from China (2010–2012) ................................................................................................................................... 20 II.3. Change in Mexico Market Share Vis-à-Vis China and Factor Intensity ........................................ 21 III.1. Selected Countries: Structure of Property Taxes in 2009 .............................................................. 37 III.2. Tax Expenditures: Estimates in Latin America ............................................................................. 38 III.3. Quantification of Options to Increase Non-Oil Tax Revenue ....................................................... 42 V.1. Labor Market Indicators ................................................................................................................. 58 Figures I.1. Trade to GDP: Mexico vs. Selected Emerging Markets .................................................................. 5 I.2. Trade with World and U.S. ............................................................................................................. 5 I.3. Annual GDP Growth Rates ............................................................................................................. 6 I.4. Workers’ Remittances vs. U.S. Value of Construction ................................................................... 7 I.5. Workers’ Remittances vs. Value of Construction ........................................................................... 7 I. 6 Financial Integration: Mexico vs. Selected Comparator ................................................................. 8 I.7 Financial Integration: Foreign Assets and Liabilities ...................................................................... 8 I.8 Foreign Liabilities Position ............................................................................................................. 8 I.9 Foreign Asset Position ..................................................................................................................... 8 I.10 Portfolio Liabilities vs. Real World Interest Rates .......................................................................... 9 I.11 Annual Change in Residents’ Portfolio Investment Abroad ......................................................... 10 I.12 Daily Foreign Exchange Market Turnover .................................................................................... 11 I.13 VIX and FX Rate ........................................................................................................................... 11 I.14 Response of Real GDP to Selected Indicators ............................................................................... 12 I.15 Response of Market Prices to Shocks in Global Financial Conditions ......................................... 13 I.16 Impulse Response to One Standard Deviation Innovations in the VIX ........................................ 14 II.1. Share of U.S. Manufacturing Imports by Origin ............................................................................ 16 II.2. Mexico’s Market Share of U.S. Imports within each Manufacturing Category ............................ 17 II.3. Auto and Other Manufactured Goods Exports .............................................................................. 18
3
II.4a. 2005–2007 Period .......................................................................................................................... 19 II.4b. 2010–2012 Period .......................................................................................................................... 19 II.5. Gains in Mexico’s Market Share Atributed to Each Competitor ................................................... 20 II.6. Real Dollar Annual Wages: Mexico and China ............................................................................. 21 II.7. Nominal Exchange Rate ................................................................................................................. 22 II.8. Mexico: Unit Labor Costs and Productivity in Manufacturing ...................................................... 22 II.9. U.S. Manufacturing-Outsourcing Cost Index ................................................................................. 22 III.1. Selected Fiscal Sector Indicators ................................................................................................... 30 III.2. Public Spending in Health and Pensions ....................................................................................... 31 III.3. Total Tax Revenue and Income Levels ......................................................................................... 31 III.4. Total Tax Revenue Collection in 2010 .......................................................................................... 32 III.5. Direct-to-Indirect Tax rations Across Country Groups ................................................................. 33 III.6. Structure of Tax Revenues in 2002 and 2010 ................................................................................ 33 III.7. Selected Countries: VAT Collection in 2010 ................................................................................ 34 III.8. Selected OECD Countries: VAT Trends ....................................................................................... 35 III.9. Selected OECD Countries: Income Taxes ..................................................................................... 36 III.10. Subnational Government Revenue and Property Taxes, 2009 ...................................................... 37 III.11. Selected Countries: Environmental Taxes in 2010 ........................................................................ 38 III.12. Income and Tax Distribution by Income Decile ........................................................................... 39 III.13. Selected Implicit Subsidies by Income Decile .............................................................................. 40 IV.1. Total Private Sector Financing ....................................................................................................... 46 IV.2a. Domestic Private Bonds ................................................................................................................. 46 IV.2b. Equity ............................................................................................................................................. 46 IV.3. Bank Credit to Private Sector ......................................................................................................... 47 IV.4. Banking Credit ............................................................................................................................... 47 V.1. Urban Unemployment .................................................................................................................... 57 V.2. Unemployment and Participation ................................................................................................... 59 V.3. Unit Labor Costs ............................................................................................................................ 59 V.4a. Migration to the U.S. (thousand per year) ...................................................................................... 60 V.4b. Migration to the U.S. (migrants per ten thousand people) ............................................................. 60 V.5. Migration and U.S. Construction .................................................................................................... 60 V.6. Labor Participation and Migration ................................................................................................. 61 Appendixes II.1. Data and Methodology Used .......................................................................................................... 27 III.1. Breakdown of Tax Expenditures in 2008, 2009, and 2010 ........................................................... 44 IV.1. Cross-Country Analysis of Factors Behind Financial System Development ................................. 50 Appendix Tables IV.1. Summary of Variables .................................................................................................................... 53 IV.2. Bank and Market-Based Financing: Basic Specification ............................................................... 54 IV.3. Bank and Market-Based Financing: Crises .................................................................................... 55 IV.4. Bank and Market-Based Financing: Institutional Factor ................................................................ 56 Appendix Figure IV.5. Bank Credit to the Private Sector ................................................................................................... 51
4
I. MEXICO: A CLOSER LOOK AT GLOBAL SPILLOVER CHANNELS1
Mexico is a highly open economy with strong real and financial links to the rest of the world. It has close linkages to the U.S. through trade and remittances, and thus is particularly sensitive to U.S. developments. In turn, Mexico’s open capital account, good macroeconomic fundamentals and liquid foreign exchange markets have led to a close integration with global financial markets, resulting in substantial portfolio flows. This warrants close vigilance to the risks of spillovers from global turbulence.
A. Introduction
1. Mexico is a highly open emerging market with strong linkages to the rest of the world. This chapter discusses and illustrates these linkages, and provides some examples of possible spillovers to Mexico of global shocks through the real and/or financial channels. Section B provides a description of Mexico’s main linkages and potential channels of transmission. Section C illustrates the relative importance of real sector and financial shocks on Mexico’s GDP, and the sensitivity of Mexico’s asset prices to global developments. Section D concludes.
B. Mexico’s Linkages with the U.S. and the Rest of the World
2. Mexico has three key direct external linkages: (i) substantial trade with the United States— particularly in manufacturing; (ii) large remittance flows from the U.S. and (iii) highly open and liquid financial markets. Mexico has significant trade ties with the U.S., and Mexican manufacturing firms are highly integrated into the U.S. supply chain.2 Mexico also receives significant remittance flows from the U.S. In addition, Mexico’s deep and liquid foreign exchange and domestic government bond markets make it susceptible to risks of financial contagion from global developments. These linkages are discussed in more detail below.
Trade channel
3. The U.S. is Mexico’s largest trading partner. Exports to the U.S. are about a quarter of GDP and account for 80 percent of Mexico’s total exports. Approximately 80 percent of total exports are in manufactured goods (95 percent of total non-oil exports). Trade openness has risen significantly over the last 30 years, particularly after Mexico joined
1 Prepared by Gilda Fernandez.
2 Trade to other regions has grown quickly in recent years, but given the gradual process of diversification, trade exposure to the U.S. is still high.
5
NAFTA in 1994. Mexico’s share in the U.S. manufacturing market temporarily declined after China joined the WTO but it has recovered since 2005.3
4. Through the trade channel, Mexico’s output fluctuations have been closely linked to that of the U.S. Several papers in the literature provide robust evidence that output fluctuations have become closely synchronized with the U.S. cycle after Mexico joined NAFTA in 1994.4 This empirical evidence includes simple correlation coefficients, variance decompositions from bivariate vector autoregressions (VARs) of Mexico’s GDP and exports with U.S. output, and results from cointegration tests between Mexico and U.S. GDP.5 For instance, bivariate VAR results show that U.S. GDP explains about 30 percent of the variability of Mexico’s real GDP. Results also show that the strongest links are related to trade and manufacturing, underscoring Mexico’s manufacturing exports integration with the
3 See “What Explains Mexico’s Recovery of U.S. Market Share?,” Chapter 2 of the Selected Issues Paper.
4 The fact that Mexico has not had home-grown crises since 1994 has probably contributed to the higher correlation with the U.S. business cycle.
5 See, inter alia, “How has NAFTA Affected the Mexican Economy: Review and Evidence,” M. Ayhan Kose, Guy M. Meredith, and Christopher M. Towe, IMF Working Paper 04/59, April 2004; “External Shocks and Business Cycle Fluctuations in Mexico: How Important are U.S. Factors?” Sebastian Sosa, IMF Working Paper 08/100, April 2008l; “A Note on Mexico and U.S. Manufacturing Industries’ Long-term Relationship,” Daniel Chiquiar and Manuel Ramos-Francia, Banco de Mexico Working Paper 2008– 08; and “Mexico’s Business Cycles and Synchronization with the USA in the Post-NAFTA Years,” William Miles and Chu-Ping C. Vijverberg, Review of Development Economics, 15(4), 638–650, 2011.
0
10
20
30
40
50
60
70
80
90
Pola
nd
Mex
ico
Chile
Chin
a
Sout
h A
fric
a
Phili
ppin
es
Turk
ey
Russ
ia
Indo
nesi
a
Indi
a
Egyp
t
Colo
mbi
a
Braz
il
Figure 1. Trade to GDP: Mexico vs. Selected Emerging Markets
Sources: IFS, Haver Analytics, country authorities, and staff calculations.
0
10
20
30
40
50
60
70
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
Total TradeUS TradeExports to USImports from US
Figure 2. Mexico: Trade with World and U.S.(In percent of GDP)
Sources: Haver Analytics, and staff calculations.
NAFTA(Jan. 1994)
China WTO Accession(Dec. 2001)
61
39
24
15
Country Share
United States 77.3Canada 2.6China 2.2Brazil 1.9Colombia 1.4Spain 1.2India 1.1Germany 1.0
1/ includes only countries with shares greater than or equal to 1 percent.
Source: EMED database and staff calculations.
Table 1. Top Mexico Export Destinations 1/ (percent share to total exports, as of December 2011)
Sample: 1994.1-2011.4 (Post-NAFTA)
Unrestricted Cointegration Rank Test (Trace)
Hypothesized Trace 0.05
No. of CE(s) Eigenvalue Statistic Critical Value Prob.**
None * 0.135 16.594 15.495 0.034
At most 1 * 0.082 6.152 3.841 0.013
Trace test indicates 2 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Table 2. Mexico and US GDP Cointegration Post-NAFTA
6
U.S. manufacturing supply chain, particularly for the automobile industry.
5. Trade ties with the U.S. are expected to remain strong. In the past 10 years, some diversification in terms of Mexico’s export markets has occurred, but it still remains limited. For example, real export growth to Europe and emerging and developing countries has been very strong over the last decade, but exports to these countries remain a small share of total exports. Mexico’s export sector is expected to remain an important engine for growth. This is supported by continued FDI into the manufacturing sector, improved relative labor costs particularly with respect to China, and relatively high oil prices over the medium term, increasing transportation costs.6
6 Announced FDI plans in Mexico through August 2012 amount to about US$6.5 billion, mainly in the automotive industry.
-8
-6
-4
-2
0
2
4
6
8
10
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
Mexico GDP
US GDP
Figure 3. Annual GDP Growth Rates(In percent)
Sources: Haver Analytics
pre-NAFTA post-NAFTA
US/Mex real GDP 0.02 0.87
US/Mex industrial production 0.23 0.65
US/Mex manufacturing 0.11 0.75
US imports/Mex exports -0.04 0.92
US/Mex consumption -0.26 0.76
US consumption/Mex exports -0.24 0.68
US/Mex investment 0.16 0.75
US/Mex construction 2/ … 0.73
Sources: Haver Analytics and staff calculations.
1/ All variables are in real terms, except construction.
2/ Sample is from 2007-2011.
Table 3. Matrix of Correlation Coefficients 1/
(year-on-year growth rates)
US variables Pre-NAFTA Post-NAFTA
Industrial Production 19.6 24.4
Manufacturing 13.9 28.1
Retail Sales 5.6 20.9
Auto Sales 7.9 22.3
Construction 1/ … 15.6
GDP 15.0 29.8
Gov. Cons. 9.7 0.2
P. Cons. 6.3 15.2
Imports 15.8 29.5
Exports 8.0 29.0
P. Inv. 17.9 23.8
Sources: Haver Analytics and staff estimates.
1/ Sample from 2006-2011 only.
Table 4. Variance Decompositions from
Bivariate VARs: Mexico Real GDP
US Variables Pre-NAFTA Post-NAFTA
Industrial Production 7.5 43.8
Manufacturing 0.8 46.4
Retail Sales 0.4 25.7
Auto Sales 19.7 33.8
Construction 1/ … 11.9
GDP 3.0 40.4
Gov. Cons. 0.2 0.6
P. Cons. 1.3 17.1
Imports 1.2 44.4
Exports 6.7 49.9
P. Inv. 0.8 36.6
Sources: Haver Analytics and staff estimates.
1/ Sample from 2006-2011 only.
Table 5. Variance Decompositions from
Bivariate VARs: Mexico Real Exports
7
Remittances
6. Remittances from the U.S., particularly from workers in the construction sector, represent an important source of income for Mexico. At about 2 percent of GDP in 2011, remittances have been highly correlated with U.S. construction activity, particularly since 2006, following the downfall of the U.S. construction sector.
EU G7 US Total
2002 7.0 -1.0 -3.9 -3.9 -3.32003 1.2 5.6 -2.3 -2.7 -2.32004 15.6 4.9 8.5 8.9 8.62005 32.8 29.6 8.0 7.3 8.52006 27.8 15.8 11.5 11.3 12.02007 25.1 25.4 2.7 1.3 3.12008 17.9 13.4 0.0 -0.5 0.52009 -27.8 -35.4 -24.0 -24.7 -24.32010 36.1 19.0 23.5 23.8 24.32011 28.7 28.0 11.2 11.3 12.0
Average real growth, 2002-2011 16.4 10.5 3.5 3.2 3.9
Real exports in percent of Mexico GDP
2011 5.0 2.6 39.8 36.9 42.4
Sources: EMED database, country authorities, and staff calculations.
Emerging and Developing
Countries Advanced Countries
Table 6. Mexico: Real Export Growth by Region
Company Amount
Ford 1,300
Audi 1,300
Ternium 1,000
General Motors 900
Honda 800
Mazda 500
Daimler Trucks 300
Jatco 200
Yorozu 70
Denso 50
Nippon Steel 40
Hennings Automotive 14
Total 6,474
Source: Newspaper articles.
Table 7. Announced Plans for Future FDI
in Mexico, 2012 (in million US dollars)
0
200000
400000
600000
800000
1000000
1200000
1400000
0
500
1000
1500
2000
2500
3000
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
RemittancesUS Construction (RHS)
Figure 4. Mexico: Workers' Remittances vs. U.S. Value of Construction (U.S. millions, SA)
Sources: Haver Analytics, and staff calculations.
correlation coefficient: +0.83
-60000
-40000
-20000
0
20000
40000
60000
80000
-250
-200
-150
-100
-50
0
50
100
150
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
RemittancesUS value of construction (RHS)
Figure 5. Mexico: Workers' Remittances vs. Value of Construction (Cyclical components)
Sources: Haver Analytics, and staff calculations.
correlation coefficient: +0.82correlation coefficient: +0.22
8
Financial Channel
7. Mexico is highly financially integrated with the rest of the world. Mexico’s stock of foreign assets and liabilities in percent of GDP, a common measure of financial integration, was about 112 percent in mid-2012. Mexico is a net recipient of foreign capital, with its foreign liabilities at 72 percent of GDP, doubling over the last 30 years.
8. The composition of Mexico’s foreign assets and liabilities has evolved. Foreign portfolio liabilities have increased rapidly since 2009, reaching more than 30 percent of GDP in mid-2012 and represent the most important source of potential financial spillovers for Mexico. Inward FDI has also increased, reaching 28.5 percent sof GDP. On the asset side, Mexico has built up reserves (broadly maintaining coverage in terms of balance sheet exposures) and increased foreign investments abroad.
9. Close to half of the FDI in Mexico comes from the U.S., mostly in the automotive industry. Another 11 percent comes from Spain, mostly banking- and tourism-related. In terms of outward FDI, the U.S. is the top destination (46.2 percent) followed by Brazil (15.5 percent).
0
50
100
150
200
250
Chile
Sout
h Af
rica
Pola
nd
Russ
ia
Phili
ppin
es
Chin
a
Mex
ico
Braz
il
Turk
ey
Colo
mbi
a
Indo
nesi
a
Egyp
t
Indi
a
Figure 6. Financial Integration: Mexico vs. Selected Comparator Countries (In percent of GDP; 2011)
Sources: IFS
20
30
40
50
60
70
80
90
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Total AssetsTotal Liabilities
Figure 7. Mexico Financial Integration: Foreign Assets and Liabilities (Percent of GDP)
Sources: Mexican Authorities, and staff calculations.
65
36
0
10
20
30
40
50
60
70
80
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Debt liabilitiesInward FDIPortfolio liabilities
Figure 8. Foreign Liabilities Position(Percent of GDP)
Sources: Banxico
0
10
20
30
40
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Reserve assets Debt assets
Outward FDI Portfolio assets
Figure 9. Foreign Asset Position(Percent of GDP)
Sources: Banxico
9
10. Since 2010, portfolio inflows have gained substantial momentum, due to both domestic and external factors. Domestic pull factors include Mexico’s strong fundamentals and sound macroeconomic management, deep and liquid exchange rate and bond markets, and highly open capital account. The inclusion of Mexico in Citigroup’s WBGI was also a significant contributor to a stock increase in portfolio investment since 2010. External factors have been dominated by the lax monetary conditions in advanced economies and investors’ consequent search for yields. Recurrent bouts of unsettled conditions in Europe have also affected external conditions for Mexico. The U.S. is also Mexico’s main source and destination of portfolio investment. Over half of portfolio liabilities are with U.S. residents and another one third are with Europe.
Country billion US$ share of total percent of GDP
United States 197 60 19.0Netherlands 37 11 3.6Spain 36 11 3.5United Kingdom 10 3 1.0Canada 8 2 0.8Germany 8 2 0.8Switzerland 5 2 0.5France 4 1 0.4Japan 3 1 0.3
Virgin Islands, British 3 1 0.3
Others 17 5 1.6
Total 330 100.0 31.9
Source: CDIS, country authorities, and staff calculations.
Table 8. Mexico: Top Ten Sources of FDI, 2010
Country billion US$ share of totalpercent of GDP
United States 48 46.3 4.7Brazil 16 15.6 1.6Netherlands 8 7.7 0.8Colombia 4 4.2 0.4Spain 4 3.6 0.4Dominican Republic 3 2.5 0.2Netherlands Antilles 2 1.9 0.2Hungary 2 1.7 0.2Chile 2 1.5 0.2Guatemala 1 0.9 0.1
Others 14 13.5 1.4
Total 104 100.0 10.0
Source: CDIS, country authorities, and staff calculations.
Table 9. Mexico: Top Ten Destinations of FDI, 2010
2007 2008 2009 2010
Advanced 96 95 97 96
US 62 59 59 54
EU 27 29 29 31
G7 78 75 75 74
LATAM and Caribbean 2 3 2 3
Sources: CPIS and staff calculations.
(% share to total)
Table 10. Mexico: Stock of Portfolio Liabilities by Region
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
-15000
-10000
-5000
0
5000
10000
15000
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Portfolio liabilities (USD, millions)Real world interest rates (in percent, RHS) 1/
Figure 10. Mexico: Portfolio Liabilities vs. Real World Interest Rates
Sources: Haver Analytics, country authorities, and staff estimates.1/ Real world interest rate estimated as the difference between LIBPR and US CPI.
Lehman collapse (0.5% of GDP)
US subprime crisis (0.9% of GDP)
10
11. Experience during the U.S. subprime crisis and the collapse of Lehmann shows that portfolio flows in Mexico can be volatile. Portfolio outflows during these periods ranged between 0.5 and 1 percent of GDP in one quarter. In turn, outflows associated with residents’ portfolio investments abroad have also been volatile and substantial, reaching about 25 percent of the stock in 2009.
12. Financial integration in Mexico is associated with liquid foreign exchange and local sovereign debt markets:
Foreign exchange market. Mexico has a deep and liquid foreign exchange market. The Mexican peso is traded globally onshore and offshore. Based on the 2010 BIS Triennial Central Bank Survey, the daily average foreign exchange market turnover in Mexico was estimated at US$17 billion (almost double the amount recorded in 1998). The bulk of these operations consisted of cross-border trades involving peso/US dollar transactions conducted by dealers abroad. The depth and liquidity of the foreign exchange market, in the context of Mexico’s floating exchange rate regime, has allowed market participants worldwide to use the peso to take positions in reaction to external developments. As such, a large part of Mexico’s currency volatility in recent years has reflected global sentiments rather than country-specific factors.
Local government debt markets. In recent years, the deepening of the foreign exchange market has been accompanied by a substantial increase in foreign investment in domestic currency-denominated government bonds, reaching 32 percent of the total stock. Following Mexico’s inclusion in the WGBI, the pool of investors in the government bond market has shifted to long-term institutional investors. During recent periods of global volatility, investors have opted to hedge foreign exchange exposures but not to divest from government paper. As a result, the exchange rate has shown significant volatility during periods of heightened global stress but domestic interest rates have remained stable.
-30
-20
-10
0
10
20
-20
-10
0
10
20
30
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Figure 11. Annual Change in Residents' Portfolio Investment Abroad (In USD, billions) 1/
Note: 1/ A positive value denotes an FX outflow in the BOP
11
C. Sensitivity of Mexico’s Growth and Asset Prices to Global Shocks
13. U.S. shocks explain a large share of Mexico’s macroeconomic fluctuations after NAFTA. To illustrate the relative importance of real and financial channels on growth, a multivariate VAR was estimated using the VIX, U.S. GDP, world interest rates, terms of trade and Mexico real GDP (with quarterly data from 1994 to 2011).7 Results from impulse response functions indicate that Mexico’s output is most affected by shocks to U.S. output. For instance, a one standard deviation increase in U.S. GDP (0.7 percentage points) would result in a 0.6 percentage point increase in Mexico’s GDP. In turn, a one standard deviation increase in the terms of trade would increase output by 0.4 percentage point. A one standard deviation increase in the VIX would result in a 0.3 percentage point decline in Mexico’s real GDP after one quarter.
7 The real world interest rate was estimated as the difference between the LIBOR and U.S. inflation.
0
2
4
6
8
10
12
14
16
18
1998 2001 2004 2007 2010
Figure 12. Daily Foreign Exchange Market Turnover(USD, billions)
Sources: BIS Triennial Central Bank Survey, December 2010.
8
9
10
11
12
13
14
15
16
-60
-40
-20
0
20
40
60
80
100
120
140
160VIX
MXN/USD Spot Rate (RHS)
Figure 13. VIX and FX Rate
Sources: Bloomberg and Haver Analytics.
Table 11. Mexico: Stock of Domestic Public Sector Debt Held by Foreigners (end-September 2012)
Outstanding(billions USD) billions USD percent of total percent of 2011 GDP
Total 332.91 108.1 32.5 9.4Short-Term (Cetes) 66.01 29.6 44.9 2.6Long Term 1/ 266.91 78.5 29.4 6.8
Sources: Country authorities and staff calculations.
1/ Other Federal Government Bonds; IPAB bonds and BREMS are not included
Held by Foreigners
Local Currency Debt(issued domestically, end-September 2012 in billions USD)
12
Figure 14. Response of real GDP to selected indicators (Multivariate VAR)
Sources: Haver Analytics and staff calculations.
14. Spikes in the volatility of Mexico’s asset prices are closely associated with periods of global stress. To examine the sensitivity of Mexico’s asset prices to global market developments, four markets were considered: the foreign exchange market, stock market, sovereign debt market, and the corporate debt market. The 3-month rolling standard deviation of asset prices in each of these markets was calculated and compared to the 3-month rolling standard deviation of the VIX, which was used as an indicator of global stress. Results show that spikes in the volatility of Mexico’s asset prices in all four markets are closely correlated with periods of global stress.
-.008
-.004
.000
.004
.008
.012
1 2 3 4 5 6 7 8 9 10 11 12
Response of DLMEXRGDP to VIX
-.008
-.004
.000
.004
.008
.012
1 2 3 4 5 6 7 8 9 10 11 12
Response of DLMEXRGDP to DLUSRGDP
-.008
-.004
.000
.004
.008
.012
1 2 3 4 5 6 7 8 9 10 11 12
Response of DLMEXRGDP to WINTRATE
-.008
-.004
.000
.004
.008
.012
1 2 3 4 5 6 7 8 9 10 11 12
Response of DLMEXRGDP to TOT
Response to Cholesky One S.D. Innovations ± 2 S.E.
13
Figure 15. Response of Market Prices to Shocks in Global Financial Conditions
Sources: Haver Analytics and staff calculations.
15. Results of bivariate VARs illustrate the sensitivity of asset prices to global stress. For example, impulse response functions show that a one standard deviation shock in the VIX is estimated to result in a 1 percent depreciation, a 4 percent decline in stock prices, an 8 percent increase in EMBI Mexico spreads (about 10 basis points), and an almost 9 percent increase in the CEMBI Mexico spreads (over 30 basis points).
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
0
5
10
15
20
25
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
VIX vs FX Rate(3-month rolling standard deviation)
VIX (LHS)
FX rate (RHS)
0
5
10
15
20
25
0
50
100
150
200
250
300
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
VIX vs EMBI+ MEX(3-month rolling standard deviations)
EMBI+ MEX (LHS)
VIX (RHS)
0
5
10
15
20
25
0
500
1000
1500
2000
2500
3000
3500
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
VIX vs IPC(3-month rolling standard deviations)
IPC
VIX
0
5
10
15
20
25
0
50
100
150
200
250
300
350
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
VIX vs CEMBI Mexico(3-month rolling standard deviation)
CEMBI MEX
VIX
14
Figure 16. Impulse Response to One Standard Deviation Innovations in the VIX 1/ (Bivariate VARs)
Sources: Haver Analytics, Bloomberg, and staff calculations. 1/ One standard deviation of the VIX = 8.4
D. Conclusion
16. Over the last two decades, Mexico has become highly globally integrated through trade and financial channels. Increased integration has been very beneficial, supporting productivity and growth, particularly in the manufacturing sector, but it has also heightened the sensitivity of output and financial markets to global developments. Thanks to its strong fundamentals and robust policy frameworks, including the flexible exchange rate and foreign exchange buffers, Mexico is well positioned to weather global volatility.
-.004
.000
.004
.008
.012
.016
1 2 3 4 5 6 7 8 9 10 11 12
Response of DLFX to CholeskyOne S.D. VIX Innovation
-.05
-.04
-.03
-.02
-.01
.00
.01
1 2 3 4 5 6 7 8 9 10 11 12
Response of DLIPC to CholeskyOne S.D. VIX Innovation
-.04
-.02
.00
.02
.04
.06
.08
.10
.12
1 2 3 4 5 6 7 8 9 10 11 12
Response of DLEMBIMEX to CholeskyOne S.D. VIX Innovation
-.02
.00
.02
.04
.06
.08
.10
.12
1 2 3 4 5 6 7 8 9 10 11 12
Response of DLCEMBIMEX to CholeskyOne S.D. VIX Innovation
15
REFERENCES Bayoumi, Tamim and Trung Bui, “Unforeseen Events Wait Lurking: Estimating Policy
Spillovers from U.S. to Foreign Asset Prices,” IMF Working Paper 11/183. Bayoumi, Tamim and Andrew Swiston, “Foreign Entanglements: Estimating the Source and
Size of Spillovers Across Industrial Countries,” IMF Working Paper 07/182, July 2007.
Benelli, Roberto and Srideep Ganguly, “Financial Linkages Between United States and Latin
America—Evidence from Daily Data, IMF Working Paper 07/262, November 2007. Chiquiar, Daniel and Manuel Ramos-Francia, “A Note on Mexico and U.S. Manufacturing
Industries’ Long-term Relationship,” Banco de Mexico Working Paper 2008–08. International Monetary Fund, 2011, “The United States Spillover Report—2011 Article IV
Consultation, IMF Country Report 11/203, July 2011. ______, 2012, “Regional Economic Outlook: Western Hemisphere: Rebuilding Strengthen
and Flexibility, April 2012. ______. 2012, “Global Financial Stability Report: Restoring Confidence and Progressing on
Reforms,” October 2012. Kose, M.A., G.M. Meredith, and C.M. Towe, “How has NAFTA Affected the Mexican
Economy: Review and Evidence,” IMF Working Paper 04/59, April 2004. Lane, P.R. and G.M. Milesi-Ferretti, “External Wealth of Nations Mark II: Revised and
Extended Estimates of Foreign Assets and Liabilities, 1970–2004,” IMF Working Paper 06/69, March 2006.
Miles, William and Chu-Ping C. Vijverberg, “Mexico’s Business Cycles and
Synchronization with the USA in the Post-NAFTA Years,” Review of Development Economics, 15(4), 638–650, 2011.
Milesi-Ferretti, G.M., F. Strobbe, and N. Tamirisa, “Bilateral Financial Linkages and Global
Imbalances: A View on the Eve of the Financial Crisis, IMF Working Paper 10/257. Sosa, Sebastian, “External Shocks and Business Cycle Fluctuations in Mexico: How
Important are U.S. Factors?” IMF Working Paper 08/100, April 2008.
16
II. WHAT EXPLAINS MEXICO’S RECOVERY OF U.S. MARKET SHARE?1
Mexico’s market share in the U.S. manufacturing market has staged a strong recovery since 2005. By mid-2012, Mexico’s share of U.S. manufactured imports reached near 15 percent, surpassing its post-NAFTA peak. In contrast with earlier trends, Mexico’s gains in the 2010–2012 period coincided with a fall of China’s market share. Recent gains in market share may have been driven in part by improved relative unit labor costs in dollars. Other factors could include the reemergence of a location advantage due to higher oil prices and a change in inventory management among U.S. firms, and a reassessment of benefits of relocating to Mexico (including the protection of proprietary technologies). As a result, global manufacturers are increasingly relocating to Mexico (near-shoring of production), particularly in the automotive, aerospace and electronics and appliances sectors.
A. Background and Recent Developments
1. Mexico’s manufacturing sector, a key engine of growth, is highly integrated with the U.S. manufacturing supply chain. Approximately 80 percent of total Mexican exports go to the U.S. (representing almost 22 percent of GDP), most of which come from the manufacturing sector. Following its entry into NAFTA, Mexico’s participation in the U.S. manufacturing market grew from less than 10 percent of U.S. manufacturing imports in 1996 to close to 15 percent in mid-2012 (Figure 1), with industries increasingly vertically-integrated with those of the U.S., particularly in the automobile sector.
2. China’s entry to the World Trade Organization (WTO) in 2001 transitorily undercut Mexico’s export share in the U.S. market. China’s low-cost manufacturing base and ample production capacity had a major impact on Mexico’s exports to the U.S. in the wake of its entry to the WTO. Between 2001 and 2005, Chinese manufacturing exports to the U.S. expanded at an average annual rate of 24 percent, while Mexico’s export growth decelerated sharply to 4 percent on average over the same period.2 As a result, China’s share of U.S. manufacturing imports almost doubled by 2005, eroding the 1 Prepared by Herman Kamil and Jeremy Zook. Enrique Flores and Esteban Vesperoni also contributed to the project.
2 The relocation of manufacturing activities (including from the Maquila industry) from North America to China, and Asia more generally, also explained the low export growth.
5
10
15
20
25
30
1996 1998 2000 2002 2004 2006 2008 2010 2012
China
Mexico
Canada
Japan
Newly Indust. Econ.
Figure 1. Share of US Manufacturing Imports by Origin(In Percent)
17
previous gains of market share by Mexico (Figure 1). China’s crowding out of Mexican exports in the U.S. market was the result of a loss of comparative advantage in several labor-intensive manufacturing sectors in which it had previously specialized, including apparel, office machinery, furniture and photographic and optical equipment sectors (Chiquiar, Fragoso and Ramos-Francia, 2008).3
3. Since 2005 Mexico has regained ground in the U.S. import market, particularly after the recent global crisis.4 Mexican manufacturing exports clawed back their share of the U.S. import market over the last seven years, rising 3 percentage points to a historically-high 14.4 percent by mid-2012. Between 2005 and 2010, Mexico’s increased presence in the U.S. import market paralleled a market share gain by China, while Japan’s and Canada’s continued to decline. Since 2010, however, Mexico’s gains in the U.S. import market coincided with a decline in China’s market participation (Figure 1).5
4. The recovery in Mexico’s market share has been driven by exports of electronics, telecommunications, and transport equipment. Since 2005, Mexico’s share of U.S. imports of transport and communications products increased steadily to 18 percent, accounting for 76 percent of total Mexican manufacturing exports in the first half of 2012. The recovery in market share of other manufacturing goods, however, took place only after 2009 (Figure 2). Analyzing trade flows at a more disaggregated level (2-digit Standard International Trade Classification) shows that Mexico’s market share gains were generalized, in 20 of the 26 manufacturing import categories, jointly accounting for 80 percent of total Mexican exports (Table 1). 6
3 Several studies have used sectoral flow data to assess the impact from Chinese exports on Mexican and other Latin American producers (Freund and Ozden, 2006; Hanson and Robertson, 2007; Lederman et al., 2008; Devlin et al., 2006; Lall et al., 2005). More recently, Iacovone, Rauch and Winders (forthcoming) provide evidence on the impact of Chinese competition on Mexican manufacturing firms between 1996 and 2004.
4 Based on revealed comparative advantage measures, Chiquiar and Ramos-Francia (2008) provide evidence that the Mexican manufacturing sector reacted to the increase in China’s competition by shifting resources towards sectors where it remained competitive. This allowed the effect of China’s entry to the WTO to be only temporary.
5 In contrast, during the period 2001-2004, Mexico lost market share in U.S. imports in 13 of the 26 sectors.
6 Among the sectors that lost market share, the most important is electrical machinery, apparatus and appliances, which accounted for 14 percent of Mexican exports in 2012, and lost 1.1 percentage points in
(continued)
5
7
9
11
13
15
17
19
1996 1998 2000 2002 2004 2006 2008 2010 2012
Machinery and Transport Equipment (SITC-7)
Textiles and Metals (SITC-6)
Other (SITC-8)
Figure 2. Mexico's Market Share of US Imports within each Manufacturing Category(In Percent)
18
5. In turn, the automotive sector has been the most important contributor to the increase in aggregate market share, explaining half of the increase between 2005 and 2012. Mexico’s market share in U.S. imports of autos, auto-parts and accessories (excluding trucks) increased almost 9 percentage points over this period (Table 1), particularly since 2009 (Figure 3). By 2012, Mexico has become the second-biggest foreign supplier of autos and auto-parts to the U.S., behind Canada, exporting over one fifth of the total U.S. imports. The automotive sector accounts for one quarter of all Mexican manufacturing exports to the U.S. The large increase in production capacity and exports has been underpinned by a continued flow of foreign direct investment into the sector. 7
market share over this period. The biggest decline occurred in the U.S. import market share of apparel and clothing products, which fell 8 percentage points from its peak of 14 percent at the beginning of the last decade.
7 Mexico was the third-largest investment destination for the automotive sector in the world, receiving US$5.6bn during 2007–10, above Japan and the BRICs, according to PROMEXICO. While the industry has been largely dominated by U.S. manufacturers in the past, more recently, companies from other countries, particularly Asia, have opened or expanded operations in Mexico. See Martin (2012) for a detailed analysis of FDI flows into Mexico’s manufacturing sector in recent years.
SITC Category
Change in US
Import Share
(2005-2012)
Share of Total
Mexican Exports
(2005)
Share of Total
Mexican Exports
(2012)68 Non-ferrous metals 9.4 1.3 3.0
78 Road vehicles 8.6 21.6 25.6
75 Office machines and automatic data-processing machines 7.4 5.7 8.6
82 Furniture, and parts thereof; bedding, mattresses, mattress
supports, cushions and similar stuffed furnishings3.7 3.5 3.2
74 General industrial machinery and equipment, n.e.s, and
machine parts, n.e.s, 3.4 5.4 6.9
81 Prefabricated buildings; sanitary, plumbing, heating and
lighting fixtures and fittings, n.e.s.3.0 1.0 0.8
87 Professional, scientific and controlling instruments and
apparatus, n.e.s.2.8 4.3 4.6
61 Leather, leather manufactures, n.e.s., and dressed furskins 2.7 0.1 0.1
76Telcommunications and sound-recording and reproducing
apparatus and equipment2.6 15.2 13.4
79 Other transport equipment 2.5 0.2 0.5
72 Machinery specialized for particular industries 2.2 1.0 1.4
71 Power-generating machinery and equipment 1.9 5.4 5.7
Table 1. Sectors with the Biggest Gain in US import Market Share: 2005-2012
30
50
70
90
110
130
150
170
30
50
70
90
110
130
150
170
2007 2008 2009 2010 2011 2012
AutomobilesOther Manufactured Goods
Figure 3. Auto and other Manufactured Goods Exports(in millions of dollars; Index June 2008=100)
19
B. Evolution of U.S. Market Share: Mexico and China
6. Mexico’s gains in U.S. market share have increasingly occurred in sectors where China reduced its share. Figure 4a and 4b show the changes in Mexico’s market share in U.S. imports (horizontal axis) against those of China (vertical axis) in each of the 2-digit sectors, for the periods 2005–2007 and 2010–2012.8 In each panel, the North-West quadrant (red) indicates those sectors in which China’s market share increased while Mexico’s fell, while the South-East quadrant (green) plots those sectors, if any, in which Mexico’s share increased and China’s fell. The size of the bubbles is proportional to each sector’s contribution to the overall change in market share in each period. During 2005–2007 all the observations lie in the Northern quadrants, indicating that there was no sector in which Mexico increased and China simultaneously decreased market share. Moreover, during this period Mexico was losing participation in several sectors in which China was gaining participation. In contrast, in 2010–2012 there are several sectors (marked green) in which Mexico’s share increased while China’s fell. In addition, the number and relative importance of sectors in which China’s share went up and Mexico’s down (i.e., the number and size of red bubbles) declined in the most recent period.9
8 We exclude the period 2008–2009 to avoid temporary effects associated to the global crisis.
9 Among the sectors in which Mexico lost and China gained market share during 2010–2012, the most relevant for Mexico was “Telecommunications and sound-recording and reproducing apparatus and equipment”, which lost 1.1 percentage points over the period. This sector accounted for 13 percent of Mexican exports in 2012.
-4
-2
0
2
4
6
8
10
12
-4 -2 0 2 4 6 8Chan
ge in
Chi
na M
arke
t Sha
re
Change in Mexico Market Share
Figure 4a. 2005-2007 Period(In Percent)
Sources: IMF staff estimates based on U.S. International Trade Commission data
-6
-4
-2
0
2
4
6
-4 -2 0 2 4 6 8
Chan
ge in
Chi
na M
arke
t Sha
re
Change in Mexico Market Share
Figure 4b. 2010-2012 Period(In Percent)
Sources: IMF staff estimates based on U.S. International Trade Commission data
20
7. During 2010–2012, 40 percent of the dollar gains in sectors where Mexico increased its market share came from sectors where China’s share fell. Using an extension of the constant market shares methodology developed in Chami Batista (2008), we calculate the fraction of Mexico’s increase in market share that can be associated with China’s reduction, controlling for changes in the shares of the other competitors.10 During 2010–2012, 40 percent of Mexico’s dollar gains in those sectors where it increased market share could be attributed to China’s reduction (Figure 5).11 Table 2 presents those sectors showing the largest fraction of market share gains accruing from China.12 In contrast, during 2005–2007, half of Mexico’s increase in market share was explained by reductions by Canada and Japan, but none coming from China.
10 The gain in market share of one exporting country that is attributed to one of its competitors is proportional to the difference between the rates of growth of the value of exports of the two countries over the period. See the Appendix for a description of the method. Calculations were performed at the 2-digit level of aggregation.
11 Part of this reduction in China’s market share may be due to a shift in China’s exports towards a different set of goods.
12 See also Oviedo (2012) for an analysis of the sectors in which Mexico gained competitiveness during the period 2010–2012.
0
5
10
15
20
25
30
35
40
45
50
Figure 5. Gains in Mexico's market share attributed to each competitor
Sources: IMF staff estimates based on data from USITC
China ChinaJapan JapanEU EUNIE NIELAC LACCanada CanadaOther Other
2010-2012 2005-2007
SITC Sector DescriptionMexico's change in
US import share
Change in share
attributed to China
Percent of total
change due to
China
85 Footwear 0.4 0.3 96.3
83 Travel goods, handbags and similar containers 0.5 0.4 84.0
77 Electrical machinery, apparatus and appliances 0.6 0.5 82.9
81 Prefabricated buildings; sanitary, plumbing, heating and
lighting fixtures 2.6 1.9 73.1
89 Miscellaneous manufactured articles 1.1 0.7 68.8
82 Furniture, and parts thereof; bedding, mattresses 2.8 1.7 59.1
61 Leather, leather manufactures, and dressed furskins 1.3 0.6 50.3
69 Manufactures of metals 0.9 0.5 50.1
75 Office machines and automatic data-processing machines 2.7 1.1 42.2
Source: IMF staff estimates based on data from USITC
Table 2. Sectors in which the largest fraction of market share gain accrued from China (2010-2012)
21
8. During 2010–2012, Mexico’s changes in market share relative to China’s were larger in sectors with higher labor-to-capital intensity. Using revealed factor intensity indices at the 5-digit SITC product level constructed by Shirotori et al (2010), a linear regression was estimated of the relative change in market share of Mexico vis-à-vis China for a given product on its degree of labor-to-capital intensity and human capital intensity.13 For 2010–2012, results show that sectors with higher labor-to-capital intensity saw a bigger relative increase in market share for Mexico compared to China (Table 3). This appears consistent with the notion that Mexico’s recent gains in market share vis-à-vis China may have been driven in part by improved relative labor costs. In contrast, over the period 2005–2007, there is no systematic relationship between relative market share changes and factor intensity. The estimated value of the constant in the model indicates that, during 2010–2012, Mexico’s increase in market share was on average higher than China’s across goods.
C. What Explains Mexico’s Increased Competitiveness?
9. Several factors could have contributed to explain Mexico’s increased competitiveness:
(i) Relative labor costs in dollar terms. Wages in the manufacturing sector in China have increased at an average annual rate of 14 percent in nominal yuan terms from 2003 to 2011, and close to 20 percent annually in dollar terms (Figure 6), given the appreciation of the yuan (Figure 7). In contrast,
13 The authors construct the indices by calculating, for each good, a weighted average of the factor abundance of the countries that export this good, where the weights are variants of Balassa’s (1965) revealed comparative advantage index.
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
2003 2004 2005 2006 2007 2008 2009 2010 2011
MexicoChina
Figure 6. Real Dollar Annual Wages: Mexico and China(In USD)
Sources: Barclays and CEIC China Database
2005-2007 2010-2012
constant -0.221 *** 0.323 ***
(-0.079) (-0.078)
log(labor-to-capital intensity) -0.013 0.033 ***
(0.012) (0.011)
log(human capital intensity) 0.142 0.028
(0.042) (0.035)
Number of observations 1762 1808
R-Square 0.005 0.012
Table 3. Change in Mexico market share vis a vis China and factor intensity
(OLS estimates at the 5-digit SITC level)
22
average wages in the Mexican manufacturing sector have remained fairly constant in dollar terms, underpinned by moderate wage growth and a depreciation of the peso (Figure 7).14 In 2003, average dollar wages in Mexico were six times higher than those in China, whereas in 2011 wages were only 40 percent higher. These developments have reduced the competitive advantage that China had as a low-cost supplier of manufacturing goods to the U.S. in the early part of 2000s.15
(ii) Productivity gains in Mexico Strong productivity increases underpinned by strong investment in the manufacturing sector in Mexico have helped lower unit labor costs and increase the competitiveness of manufacturing production (Figure 8). (iii) Oil prices and transportation costs The benefits of proximity to the U.S., Mexico's key location advantage, have increased with the rise in oil prices and changes in U.S. companies’ inventory management. The price of oil increased from US$25 per barrel in early 2000s to over US$100 in October 2012, increasing substantially transoceanic freight costs. This has given Mexico a competitive edge, particularly when it comes to heavy and bulky items. Proximity, as a proxy for speed-to-market, has also gained importance 14 Subdued wage growth is in part associated with an increase in labor participation and in the growth rate of the working age population, from a reduction in migration to the U.S. since the mid-2000s (sees Selected Issues Paper “Migration and Labor Markets”).
15 Reliable data on unit labor costs in the Chinese manufacturing sector is not available, preventing an assessment of the evolution of wages in China accounting for changes in productivity.
70
90
110
130
150
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
ChinaMexico
Figure 7. Nominal Exchange Rate(Domestic currency per USD; index, 2002=100)
Source: WEO
50
60
70
80
90
100
60
70
80
90
100
2005 2006 2007 2008 2009 2010
China Mexico India Vietnam
Figure 9. US Manufacturing-Outsourcing Cost Index(Landed costs in US relative to US domestic manufacturing cost, in percent)
Source: AlixPartners
86
92
98
104
110
2007 2008 2009 2010 2011 2012
Unit Labor Costs (pesos)Productivity
Figure 8. Mexico: Unit Labor Costs and Productivity in Manufacturing (Index, 2008=100)
Sources: INEGI
23
as U.S. companies increasingly adopted outsourcing of inputs and just-in-time manufacturing. According to the 2011 U.S. Manufacturing-Outsourcing Cost Index, goods produced in Mexico had the lowest landed costs (i.e., the price at a California shipping port) for U.S. importers in 2010 compared to other key low-cost countries (Figure 9).16 This has coincided with the trend to “near-shoring” (as opposed to “off-shoring”) exploiting the advantage of a proximate manufacturing hub to the U.S. (iv) Protection of proprietary technologies The strong commitment to the protection of proprietary technologies has also helped Mexico in the relocation of FDI. Mexico has a strong reputation for protection of international intellectual property, patent and trademark rights and is a party to several international treaties, including the World Intellectual Property Organization. This has helped minimize the risk of piracy, counterfeiting and other intellectual property infringements, which is especially important in high-technology sectors as well as sectors with technologies that could be used in military applications. In January 2012 Mexico joined the Wassenaar Arrangement on Export Controls for Conventional Arms and Dual-Use Goods and Technologies.17 Joining the Waasenaar Arrangement opened up new possibilities for American and European firms to invest in Mexico in the high-tech sectors, including semiconductors, software, aerospace, lasers, sensors and chemical production. (v) Openness and commitment to free trade
Mexico’s manufacturing base has been buttressed by the economy’s openness. Mexico has one of the largest trade agreements networks, with free trade or preferential agreements with 44 countries.18 It has also shown a strong commitment to free trade, avoiding the use of trade restrictions and providing assurance to companies operating in Mexico of unrestricted access to markets and intermediate inputs. Moreover, Mexico is signatory to international standards and quality agreements facilitating the insertion of local manufacturing companies into the global supply-chains.19
16 Several cost drivers are factored into the analysis, including wages and productivity, exchange rates, and freight fees, overhead, raw materials, duties, and in-transit inventory (AlixPartners, 2011).
17 Under the Waasenaar Arrangement, the 41 signatory nations cooperate and adhere to export controls for conventional arms and dual-use goods and technologies.
18 On June 2012, Mexico was formally invited to join the ongoing negotiations for the Trans-Pacific Partnership (TPP), a proposed free trade agreement involving the United States and eight other countries. 19 For instance, Mexico is a signatory of the Bilateral Aviation Safety Agreement (BASA), which brings it into the global aerospace supply-chain by allowing aircraft parts to be shipped overseas for assembly without first undergoing an international inspection. This has bolstered Mexico’s manufacturing capabilities in the aerospace
(continued)
24
10. Several of the factors that have contributed to explain Mexico’s increased competitiveness and recouping of market share in the U.S. are likely to be structural. The locational advantage, improved unit labor costs from enhanced manufacturing productivity and increased labor participation, and trade openness have likely underpinned Mexico’s improved competitiveness in the U.S. market in recent years. Structural reform efforts to boost productivity and investment would help sustain the dynamism of manufacturing exports and boost potential GDP growth. These efforts would include, inter alia, measures to further foster competition and labor flexibility, improve education and reinforce domestic security.
sector: more than 260 aerospace companies now operate in Mexico (mostly in the Tijuana-Mexicali cluster), exporting some $4.3 billion in aircraft and parts in 2011.
25
REFERENCES AlixPartners, 2011, “U.S. Manufacturing-Outsourcing Cost Index”. Balassa, B., 1965, “Traded Liberalization and "Revealed" Comparative Advantage”, The
Manchester School of Economic and Social Studies, 33, 99–123. Chami Batista, J., 2008, “Competition between Brazil and other Exporting Countries in the
U.S. Import Market: a New Extension of Constant-Market-Shares Analysis”, Applied Economics, 40, pp. 2477–2487.
Chiquiar, D. and M. Ramos Francia, 2009, "A Note on Mexico and U.S. Manufacturing
Industries’ Long-term Relationship," Banco de México Working Paper 2008–08. Chiquiar, D. and M. Ramos Francia, 2009, "Competitiveness and Growth of the Mexican
Economy," Banco de México Working Paper 2009–11. Chiquiar, D., E. Fragoso and M. Ramos Francia, 2007, “La Ventaja Comparativa y el
Desempeño de las Exportaciones Manufactureras Mexicanas en el Periodo 1996–2005”. Banco de México Working Paper 2007–12.
Devlin, R., A. Estevadeordal and A. Rodriguez-Clare (eds.), 2006, “The Emergence of
China: Opportunities and Challenges for Latin America and the Caribbean”, Washington, DC: Inter-American Development Bank.
Freund, Caroline, and Caglar Ozden, 2007, “The Effect of China's Exports on Latin
American Trade with the World”, Mimeo, World Bank.
Hanson, G., and. R Robertson (2007). “China and the Manufacturing Exports of Other Developing Countries”. Forthcoming in R. Feenstra and S. Wei (eds.), China’s Growing Role in World Trade. Chicago: University of Chicago Press.
Iacovone, L., F. Rauch, F. and L. A. Winters, “Trade as an Engine of Creative Destruction:
Mexican experience with Chinese competition”, Journal of International Economics (forthcoming).
Iranzo, S. A. Ma, 2006, “The Effect of China on Mexico U.S. trade: Undoing NAFTA?
Report: University of California, San Diego. Lall, S., J. Weiss and H. Oikawa, 2005, “China’s Competitive Threat to Latin America: an
Analysis for 1990–2002”, Oxford Development Studies, 33(2): 163–194.
26
Lederman, D., M. Olarreaga and E. Rubiano, 2008, “Trade Specialization in Latin America: the Impact of China and India”, Review of World Economics. 144(2): 248–271.
Martin, S., 2012, “Manufacturing Magnet Lures FDI”, HSBC Global Research. Mesquita Moreira, M., 2007, “Fear of China: Is There a Future for Manufacturing in Latin
America?”, World Development, 35(3): 355–376. Oviedo, M., 2012, “The Productivity Revolution and the China (Manufacturing) Syndrome”.
Barclays Emerging Market Research. Shirotori, M., B. Tumurchudur and O. Cadot, 2010, “Revealed Factor Intensity Indices at the
Product Level”, Policy Issues Series No. 44, UNCTAD.
27
APPENDIX 1. DATA AND METHODOLOGY USED
A.1. Data
The empirical analysis is based on data on the customs dollar value of U.S. manufacturing imports (categories 6, 7, and 8 of the Standard International Trade Classification) by country of origin disaggregated at the 5-digit level. Data is at annual frequency from 1996 through June 2012. The source is the United States International Trade Commission (USITC).
A.2. Competition Between Mexico and China in the U.S. Manufacturing Import Market
Below we describe an extension of the Constant Market Share (CMS) analysis derived by Chami Batista (2008), attributing the gains or losses in import market shares by a given country to each of its competing exporters. The ultimate goal is to determine what fraction of Mexico’s dollar value gains in market share in manufacture products came at the expense of China (vis a vis other countries that were also changing their share in the U.S. market over this period). Assume that there are n exporters to the U.S. market K, so that:
(1)
are the macro shares of each exporting country H=1,….,n. The change in the micro market share of product i between time t and time t+1 of exporter H can be defined as:
∆ , , ,,
,
,
, (2)
Dropping the subscript i for ease of notation, ∆ , may be defined as the part of the change in the micro share of exporter H that can be ascribed to the change in the micro share of exporter J, such that:
∆ ∆ , (3)
bearing in mind that:
28
0 (4)
That is, when exporters compete for a specific market, the changes in a given country’s share of the market necessarily come at the expense of all the others’ shares together. Starting from (4), it can be shown that the change in market share of exporter H for product i that can be attributed to exporter J is equal to:
∆ , , ∆ ,,
, ∆ ,
,
, (5)
or
∆ , , ∆ , , ∆ , , (6) which is equal to the difference between the changes in market shares for a product i between H and J, where each term is weighted by the market share at initial time t of the other competitor. Note that the only case where the increase country H’s import share to the U.S. will be unambiguously equal to the decrease in market share by country J is when country’s J share of the market for product i was 100 percent in period t (and consequently the share of country H was 0 in the initial period). Chami Batista (2008) shows that the formulation in (6) satisfies four desirable properties. First, country H cannot lose or gain from itself. Second, a gain for exporter H from exporter J is equal to the loss of exporter J to exporter H. Third, the sum of the changes in market shares of one exporting country that are attributed to each of its competitors is equal to the overall change in market shares of the country. Fourth, the change in market share of one exporting country that is attributed to one of its competitors is directly related to the difference between the rates of growth of the value of exports of the two countries. It can be shown that (9) can be re-expressed as:
∆ , ,, ,
, , (7)
where , and , is the gross rate of growth in the value of exports of products i by country H and J, respectively, and m is the gross rate of growth in the value of imports of product i by country K.
29
III. THE CASE FOR TAX REVENUE MOBILIZATION IN MEXICO1
A. Introduction
1. Mexico’s current fiscal framework is a sound basis for fiscal sustainability, but over the medium term it will require measures to offset emerging challenges. Two of the most important challenges are: (i) a decline in oil revenues as a share of GDP as the economy grows (even with oil production increasing to its 2004 peak over the medium term and with oil prices remaining at its current levels); and (ii) the projected increase in health- and pensions-related spending due to aging. More generally, Mexico also needs to further rebuild fiscal buffers as a protection against possible negative shocks from the fragile global economic environment.
2. The need to anticipate and overcome these challenges creates a strong case for additional domestic revenue mobilization. The case appears even stronger when looked at from an international perspective: Mexico’s low levels of non-oil tax collection compared with OECD countries and regional peers mean that there is space for revenue mobilization to avoid compressing public spending excessively, including public investment and social spending.2
3. This paper summarizes the nature of these challenges and examines the characteristics of Mexico’s tax system from an international perspective. It highlights areas in which the country lags behind comparator countries in terms of tax yields, looks at the distributional impact of the taxes, and quantifies the potential gains from some options to boost tax collection under different scenarios.
B. Macroeconomic Considerations for Revenue Mobilization
4. Following Mexico’s fiscal stimulus in 2009, gradual fiscal consolidation efforts have partly reversed the change in the fiscal position but are yet to restore previous levels of fiscal buffers. Fiscal adjustment since 2010 has helped stabilize the gross public debt ratio at about 43 percent of GDP, but debt has yet to come down to pre-crisis levels (about 37 percent of GDP). Stepping up consolidation efforts to return to Mexico’s primary surplus before the global crisis, would put the debt ratio on a more sustained downward path (which is particularly important in the present juncture of heightened global risk). But rebuilding fiscal buffers, which in the wake of a crisis can be a lengthy process, will be complicated by the need to cope with the projected substantial medium-term decline in oil revenues and the significant upward pressure on health- and pensions- related spending. Pressures on revenues and expenditure, projected at 4–5 percent of GDP over the next
1 Prepared by Santiago Acosta-Ormaechea.
2 Non-oil tax collection in Mexico was about 10 percent of GDP in 2011.
30
decades, appear manageable, but revenue mobilization efforts would be necessary in order to avoid an excessive compression in other expenditures.
5. Oil revenue is projected to decline by 2–3 percent of GDP through 2030, from its current level of 7.6 percent of GDP. Even if production were to return to its 2004 peak by 2030 (3.4bn barrels per day) and the real price of oil were to remain constant at its 2011 level, oil revenue would drop by about 2.1 percent of GDP over the period.3 Instead, if production were to remain constant at current levels (2.5bn barrels per day), the contraction in oil revenues would be worse—about 2.8 percent of GDP. At worst, if production stays stagnant and oil prices decline as projected by the IMF’s World Economic Outlook through 2017 (remaining constant thereafter), by 2030 oil revenue would fall by 3.3 percent of GDP.4
6. Health- and pensions-related spending are projected to increase by over 2 percent of GDP through 2030.5 The increase in pension costs (about 1.2 percent of GDP) will be driven mostly by population aging (assuming no offsetting factors such as an increase in the retirement age or further reforms to the system). Health care spending is projected to rise by 1 percent of GDP over the same period, also due to aging (which explains about 60 percent of the additional spending) and by the fact that health costs grow faster than GDP as technological medical advances are adopted. Mexico’s projected increase in spending—
3 Projecting oil revenues boils down to assessing the likely paths for oil production and international oil prices. These projections keep real GDP growth at its potential rate, the real exchange rate constant and other key parameters roughly unchanged.
4 The estimation approach considered here largely follows Lopez-Murphy (2011).
5 See Escolano et al (2012). These figures refer to the consolidated general government.
Source: SHCP and IMF staff calculations.
37.843.1 44.5 42.9 43.8 43.1
1.2 1.1
4.74.3
3.4
2.6
0
1
2
3
4
5
6
0
10
20
30
40
50
60
2007 2008 2009 2010 2011 2012
Gross public debt (left)
Fiscal deficit (right)
Deficit, debt and financing needs
0
2
4
6
8
10
0
2
4
6
8
10
2011 2015 2019 2023 2027
Constant real oil price (growing production)Constant real oil price (stagnant production)Constant nominal oil price (stagnant production)
Oil sector revenue projections
Figure 1. Mexico: Selected Fiscal Sector Indicators(in percent of GDP)
31
both for pensions and health care—is in line with expected medium-term trends in other emerging-market economies (Figure 2).
C. Mexico’s Tax System: A Cross-Country Perspective
7. Revenue collection in Mexico is low by regional and OECD standards. Although differences in accounting practices across countries make precise comparisons difficult, Mexico’s tax revenue appears as an outlier taking into account income levels and collection ratios of comparator countries (Figure 3).6 At the general government level, Mexico’s total tax revenue in 2010 was not much more than half of the OECD average (18 percent of GDP compared with 34 percent).7 It was also significantly lower than in other LAC5 economies.8
6 Mexico’s measured taxes depart from the GFSM 2001 standard in that some revenues are shown net of subsidies (for instance, gasoline and electricity tariffs).
7 The OECD Revenue Statistics Database, oil-related tax revenue is included in Mexico’s total tax revenue.
8 The LAC5 group refers to the 5 largest Latin-American economies: Argentina, Brazil, Chile, Colombia, and Mexico.
Figure 2. Public Spending in Health and Pensions(in percent of GDP)
Sources: Escolano et al (2012) and staff calculations.
2.6
3.73.3
4.3
1
2
3
4
5
1
2
3
4
5
2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030
Pensions Health
Mexico: health and pensions spending
-4
-2
0
2
4
6
-4
-2
0
2
4
6
8
TurkeyRussiaChinaSaudi ArabiaAr gentinaBrazilRom
aniaM
alaysiaU
kraineLithuaniaM
exicoLatviaSouth AfricaThailandPhili ppinesIndonesiaPakistanIndiaBulgariaPolandChile
Pensions Health
Selected EMEs: change in health and pensions spending, 2010-2030
USA
GBR
AUTBEL
DNK
FRA
DEU
ITA
LUXNLD
NOR
SWE
CHECAN
JPN
FIN
GRC
ISL
IRLPRT
ESP
TUR
AUS
NZL
ARG
BRA
CHL
COL MEX
ISR
KOR
CZE
SVK
HUNSVN
POL
MEX non-oil
10
15
20
25
30
35
40
45
50
10
15
20
25
30
35
40
45
50
9.0 9.5 10.0 10.5 11.0 11.5Ta
x re
venu
e to
GD
P ra
tio (i
n pe
rcen
t)
Tax
reve
nue
to G
DP
ratio
(in
perc
ent)
Log of GDP per capita (PPP prices)
Figure 3. Total Tax Revenue and Income Levels(OECD and Argentina, Brazil and Colombia; 2006-10 unweighted avg)
Note: Data refer to the general government level.Source: IMF staff calculations based on OECD revenue database and WEO.
32
If oil revenues are excluded, the difference with comparator countries is even more pronounced.
8. Mexico’s low revenue ratio is explained by several factors. Some of the difference is due to design choices in tax structures or in public institutions. For instance, taxes on income, profits and capital gains are lower because the PIT and CIT are fully integrated (with full imputation of dividend tax at the personal level). While this reduces the total take of taxes on profits and capital, it has good efficiency properties (low effective taxes on capital). Likewise, unlike countries with traditional pay-as-you-go public pension systems, Mexico does not include employer and employee contributions to pensions in the fiscal accounts—since these are recognized as being part of individual saving. However, Mexico’s revenue is also low because of: a relatively low tax compliance rate (in part due to informality in the economy); weaknesses in tax administration; a very low tax collection at the subnational level; and a somewhat narrow tax base (owing to differentiated rates and deductions in key taxes coexisting with special tax regimes) (OECD, 2011).9
9. Amid low income taxes and social security contributions, the tax structure in Mexico depends significantly on indirect taxation. While tax collection on goods and services in Mexico is broadly in line with OECD countries, income taxes (and social security contributions) represent a substantially lower share of total revenue (Figure 4). This makes overall tax collection in Mexico particularly reliant on indirect taxation compared with OECD and G7 countries—a characteristic also shared by the rest of the LAC5 (Figure 5).10
9 For tax collection at the central government level, differences are less pronounced: Mexico collected about 17 percent of GDP in 2009, whereas the OECD average was about 25 percent of GDP.
10 Direct taxes in Figure 5 include: taxes on income, profits and capital gains, social security contributions, taxes on payroll and workforce and taxes on property. Indirect taxes, in turn, include: taxes on goods and services and other taxes.
Figure 4. Total Tax Revenue Collection in 2010 1/ (selected countries; in percent of GDP)
1/ General government level. Data on Argentina, Brazil, Colombia, Poland and OECD average refer to 2009.Sources: OECD, Revenue Statistics Database and Latin American revenue statistics.
0
10
20
30
40
50
0
10
20
30
40
50
MEX MEX non-oil
CHL BRA ARG COL USA KOR POL OECD
Property Goods and servicesSSCs and payroll taxes Income, profits and capital gainsOther taxes
Tax Structure
33
10. The share of oil revenues has increased significantly since 2002, due to the increase in international oil prices (Figure 6). Despite growing diversification in Mexico’s economy, dependence of the public finances on oil production has not declined, missing an opportunity to diversify sources of revenue in line with the rest of the economy.11 Due to the significant increase in fuel subsidies, the share of excises declined sharply over the period, as these are accounted for as a negative excise.12 Property taxes, which are collected at the subnational level,
11 Despite the changes in the structure of the economy, the contribution of income taxes (PIT and CIT) has remained broadly unchanged in this period, accounting for about one fourth of total revenue. Similarly, VAT continues to account for about one-fifth of total tax revenue.
12 In 2002 fuel excises generated revenue of about 1.6 percent of GDP, whereas in 2010 they reached -0.6 percent of GDP (negative because of the net impact of the fuel subsidy). Over the period 1980–2005, fuel excises were on average about 1 percent of GDP.
Figure 5. Direct-to-Indirect Tax Ratios Across Country Groups
Note: data refer to the general government level.Source: IMF staff calculations based on OECD revenue statistics database.
0
5
10
15
20
25
30
35
40
0
5
10
15
20
25
30
35
40
Mexico LAC5 G7 OECD
in p
erce
nt o
f GD
P
in p
erce
nt o
f GD
P
Indirect taxes
Direct taxes
0
1
2
3
4
0
1
2
3
4
Mexico LAC5 G7 OECD
Direct-to-indirect tax ratio
Figure 6. Mexico: Structure of Tax Revenues in 2002 and 2010(in percent of total tax revenues)
Source: OECD, Revenue Statistics Database.
VAT21.3%
PIT13.2%
Social Security Contr.16.0%
CIT10.4%
Property1.7%
Excises3.6%
Customs duties1.1%
Oil revenue27.1%
Other5.6%
2010
VAT19.3%
PIT15.6%
Social Security Contr.17.9%
CIT10.7%
Property1.7%
Excises12.6%
Customs duties2.5%
Oil revenue13.4%
Other6.3%
2002
34
contribute less than 2 percent of total revenue—a share that has remained broadly constant over time.
D. Overview of Mexico’s Tax System
Value Added Tax (VAT)
11. Mexico’s VAT collection is below OECD and regional standards. VAT revenue, at 3.9 percent of GDP in 2010, stood significantly below the OECD and LAC5 average collection of around 6.7 percent of GDP.13 However, Mexico’s VAT statutory rate, currently at 16 percent, is not significantly below the OECD’s (currently at around 18 percent). Instead, the underperformance of the VAT in Mexico is mostly associated with exemptions and reduced VAT rates which, combined with the informality in the economy, have substantially narrowed the tax base (OECD, 2011).14 The reduced VAT rate at Mexico’s borders, unique among LAC5 or OECD countries, has also contributed to erode the overall VAT base.15 These special regimes have been reflected in the low VAT revenue ratio (actual collection over theoretical revenues derived from applying the standard rate to all final consumption), which is the lowest among OECD countries.16
13 Mexico’s VAT collection was 3.7 percent of GDP in 2011.
14 The VAT has zero rates for food and medicines and exemptions for services such as education and health.
15 Mexico has a differential border rate of 11 percent, which according to official estimates entails a tax expenditure of about 0.13 percent of GDP.
16 This measure of the effectiveness of the VAT is also referred to as VAT C-efficiency ratio.
Figure 7. Selected Countries: VAT collection in 2010 1/(in percent of GDP)
1/ Data for Argentina, Australia, Brazil, Colombia , Greece, Ireland, Netherlands, Poland, and OECD average refer to 2009. Sources: OECD, Revenue Statistics Database and Latin American revenue statistics.
0
2
4
6
8
10
12
0
2
4
6
8
10
12
JPN
CAN
MEX
AUS
CHE
ESP
KOR
ITA
GBR
COL
LUX
GRC IR
LO
ECD
FRA
BEL
NLD PR
TPO
LAR
GD
EU BRA
CHL
NO
RIS
LAU
TFI
NN
ZLSW
ED
NK
35
Income taxes
12. Despite statutory rates broadly in line with comparator countries, Mexico’s income tax revenues are low. The combined income tax revenue in Mexico (PIT and CIT), at 5 percent of GDP in 2009, is significantly below OECD levels—although, as noted, some of this can be explained by the full integration of the PIT and CIT.17 Statutory rates for the PIT and CIT, at 30 percent, are broadly in line with comparator countries.18 Mexico’s low income tax revenue collection reflects a combination of special regimes, widespread deductions and low compliance (OECD, 2011).
17 In addition, the OECD’s revenue statistics database defines two categories for taxes on income (profits and capital gains of individuals and corporates) which exclude other income taxes (such as Mexico’s recently-implemented Impuesto Empresarial a Tasa Unica (IETU)) that cannot be directly allocated to either of them. This latter tax yielded about 0.4 percent of GDP in 2009. However, even after taking into consideration these caveats, Mexico shows a particularly low yield for income taxes. See SHCP (2011b) for details.
18 In 2011 the average top PIT statutory rates in the LAC5 and OECD were 31 and 42 percent, respectively. Average top CIT rates for LAC5 and OECD were 30 and 26 percent, respectively. See Sabaini and Jiménez (2011) for further discussions on the characteristics of income taxes in Latin-American countries.
Figure 8. Selected OECD Countries: VAT Trends(in percent)
Source: OECD, 2010, Consumption Tax Trends.
0
5
10
15
20
25
30
0
5
10
15
20
25
30CA
NJP
NCH
EAU
SKO
RN
ZLLU
XM
EX ESP
GBR TU
RO
ECD
CHL
DEU
NLD FRA
AUT
ITA
PRT
BEL
IRL
FIN
POL
DN
KN
OR
SWE
ISL
VAT standard rates
2010 2000
0
0.2
0.4
0.6
0.8
1
1.2
0
0.2
0.4
0.6
0.8
1
1.2
MEX ITA
ESP
GBR
AUS
BEL
FRA
POL
PRT
ISL
DEU IR
LN
OR
FIN
SWE
OEC
DCZ
EN
LDAU
TD
NK
KOR
JPN
CAN
CHL
CHE
LUX
NZL
VAT Revenue Ratio
2008 2000
36
Subnational revenues and property taxes
13. Mexico’s revenue mobilization at the subnational level is very low compared with OECD and regional peers. At end-2009 subnational tax revenue collection reached 0.7 percent of GDP, about the lowest among OECD and LAC5 economies (Figure 10). Inadequate incentives, associated with the high reliance on transfers from the central government, and the high turnover of elected local authorities have underlain Mexico’s subpar revenue performance at the subnational level.19 This is seen especially in the low yield of property taxes in Mexico (which represent a substantial share of local taxation in most other countries), a phenomenon also explained by outdated property values. Other subnational tax bases could also be strengthened: small taxpayers (REPECOS), the local gasoline tax (administered by states), and the local vehicle-ownership tax, among others.
19 Transfers from the central government currently represent more than 90 percent of subnational level revenues.
Figure 9. Selected OECD Countries: Income Taxes
0
10
20
30
40
50
60
70
0
10
20
30
40
50
60
70
MEX
CHL
USA
CAN
KOR
POL
G7
OEC
D
Top statutory personal income tax rate 1/(in percent of gross wage earnings)
2011 2000
0
3
6
9
12
15
0
3
6
9
12
15
2002 2003 2004 2005 2006 2007 2008 2009 2010
PIT Revenue (in percent of GDP)
0
10
20
30
40
50
60
0
10
20
30
40
50
60
2002 2004 2006 2008 2010
Mexico Chile United States Canada Korea Poland G7 OECD
CIT Statutory Rate 2/
0
1
2
3
4
5
6
7
8
9
10
0
1
2
3
4
5
6
7
8
9
10
2002 2003 2004 2005 2006 2007 2008 2009 2010
CIT Revenue (in percent of GDP)
1/ For a single person without dependants based on the earnings level where the top statutory personal income tax rate first applies.2/ Basic combined central and sub-central (statutory) corporate income tax rate given by the adjusted central government rate plus the sub-central rate. Sources: OECD, Revenue Statistics, Tax Database, Corporate and Capital Income Taxes and Taxing Wages Databases.
37
Environmental taxes
14. Environmental taxes in Mexico are an outlier in international comparison. In 2010 these taxes generated negative net revenues due to Mexico’s fuel subsidies (-0.3 percent of GDP).20 Other environmentally-related taxes—such as those on pollution, transport or energy excluding fuels—have also shown a relatively weak performance from a cross-country perspective. This becomes apparent when considering the 2.6 percent of GDP collected by the EU27 in environmental taxes on average, with fuel taxes representing 1.6 percent of GDP (Figure 11).
20 The average revenue from fuel excises between 1980 and 2005 was 1 percent of GDP. However, in recent years, higher subsidies to the consumption of gasoline and diesel led to a net negative collection of these excises (0.6 percent of GDP in 2010).
0
4
8
12
16
20
0
4
8
12
16
20
GRC
MEX
IRLCH
LN
LDG
BRAU
TN
ZLCO
LARGPO
LKO
RBELAU
SO
ECDFRAN
OR
ITAJPNBRAU
SAISLESPFIND
EUCH
ED
NK
CANSW
E
State and local governments
0
1
2
3
4
5
0
1
2
3
4
5
MEX
AUT
DEU CH
LFI
NBR
ASW
EPO
LG
RCN
OR
NLD
COL
IRL
OEC
DD
NK
ESP
NZL IS
LCH
EAU
SIT
AJP
NBE
LAR
GKO
RU
SA FRA
CAN
Property taxes
Figure 10. Subnational Government Revenue and Property Taxes, 2009(in percent of GDP)
Sources: OECD, Revenue Statistics Database and Latin American revenue statistics.
MEX 1/ CHL BRA ARG COL USA CAN POL KOR OECD
Recurrent taxes on immovable property 0.2 0.6 0.4 0.4 0.5 3.1 3.1 1.2 0.8 1.1Estate and inheritance taxes 0.0 0.2 0.0 0.0 0.0 0.2 0.0 0.0 0.2 0.1Recurrent taxes on net wealth 0.0 0.0 0.0 0.4 0.4 0.0 0.1 0.0 0.0 0.1Taxes on financial and capital transactions 0.1 0.1 0.1 2.2 0.6 0.0 0.2 0.0 1.9 0.4
Other taxes on property 0.0 0.0 0.5 0.0 0.1 0.0 0.2 0.0 0.0 0.1
Total 0.3 0.9 1.1 2.9 1.6 3.3 3.6 1.2 3.0 1.8
1/ excludes impuesto a los depósitos en efectivo (IDE) in the case of taxes on financial and capital transactions.
Source: OECD, Revenue Statistics Database.
Table 1. Selected Countries: Structure of Property Taxes in 2009(in percent of GDP)
38
Tax expenditures
15. Mexico’s tax expenditures appear large, representing about 2.8 percent of GDP in 2009.21 These expenditures are associated with income taxes (1.3 percent of GDP), VAT (1.3 percent) and other taxes and tax reliefs (0.2 percent). International comparison, with all the caveats needed due to differences in methodologies to estimate them, show that Mexico’s tax expenditures appear relatively high in recent years (Table 2). These tend to also be poorly targeted while adding significant complexities and inefficiencies to the system.22
21 The estimate in 2008 and 2009 excludes tax expenditures and collection of the IETU (bringing it down by about 1 percent of GDP). It does not include estimates of tax expenditures associated with the Maquiladora sector. See Annex for a further breakdown of Mexico’s tax expenditures, including official figures for 2010.
22 See also OECD (2010a; 2010b) for further discussions on tax expenditures in OECD countries.
-1
-0.5
0
0.5
1
1.5
2
2.5
3
3.5
4
-1
-0.5
0
0.5
1
1.5
2
2.5
3
3.5
4
MEX ESP FRA LTU SVK BEL ROM DEU GRC CZE IRL LVA LUX AUT PRT ITA POL EU27
Energy excluding fuels Transport Pollution Fuels Environmental taxes in 2000
Figure 11. Selected Countries: Environmental Taxes in 2010 (in percent of GDP)
Source: European Commission, Taxation Trends in the European Union; and OECD/EEA database on environmentally related instruments.
2005 2006 2007 2008 1/ 2009 1/ Average 2005-9
Argentina 2.2 2.1 2.2 2.1 2.1 2.1Brazil 1.7 2.0 2.3 2.8 3.2 2.4Chile 4.4 4.1 4.9 4.0 4.0 4.2Colombia 3.7 4.0 3.5 - - 3.7Mexico 6.7 5.6 5.9 5.2 2.8 5.2Peru 2.1 2.2 2.2 2.1 1.8 2.1
1/ Mexico's estimate in 2008 and 2009 excludes tax expenditures and collection of the IETU (see Annex).
Sources: Secretaría de Hacienda y Crédito Público (SHCP): Presupuesto de Gastos Fiscales , various issues;
and Villela et al (2010).
Table 2. Tax Expenditures: Estimates in Latin America (in percent of GDP)
39
16. Tax expenditures associated with the export assembly sector (Maquiladoras) are significant and lead to important inefficiencies in the tax system. This regime currently provides substantial benefits to firms in terms of VAT, income taxes, and import tariffs, involves substantial fiscal costs, opens windows for tax planning and, over time, it has moved away from its initial objectives. To qualify as a Maquiladora, a firm currently needs to sell abroad only 10 percent of total sales; moreover, rules limiting eligible firms to border areas have been relaxed. This implies that in practice a large share of firms receiving these benefits produce goods and services that are sold in the domestic market.23
Distributional aspects
17. The distributions of income and tax collection in Mexico across households’ income deciles are broadly aligned. According to the latest National Survey of Household Income and Expenditure (SHCP, 2011a), the highest three deciles of the income distribution obtain 67 percent of gross national income and contribute with about 70 percent of total taxes (82 percent for income taxes). The three lowest deciles of the distribution instead receive 6 percent of gross income and contribute 5 percent of total taxes (8 percent for the VAT).
18. Explicit and implicit subsidies in Mexico tend to benefit households at the higher end of the income distribution. The regressive structure of subsidies is particularly apparent in the case of fuel subsidies and residential subsidies for electricity. VAT exemptions and reduced rates also tend to be regressive, although less so in the case of zero-rated VAT
23 Tax expenditure associated with the special income tax regime for Maquiladoras in 2010 was about 0.11 percent of GDP. This figure only reflects partially the associated fiscal cost, as it excludes the forgone revenue due to the elimination of VAT payments for ‘so-called’ temporary imports of intermediate inputs and other tax expenditures associated with the sector which are difficult to quantify.
-10
0
10
20
30
40
50
60
-10
0
10
20
30
40
50
60
I II III IV V VI VII VIII IX X
Households' income decile
Income distribution VAT collectionISR collection Total tax collection
Figure 12. Income and Tax Distribution by Income Decile(in percent)
Sources: based on SHCP (2011a)
40
goods. In all, tax expenditures in Mexico do not appear to be effective to help redistribute income to the most vulnerable groups.24
19. Mexico’s conditional cash transfers program Progresa/Oportunidades has been particularly effective in improving the after-transfers distribution of income. Recent estimations suggest that the program is a good example of redistributive efficiency, in the sense that it spends 0.36 percent of GDP and helps explain approximately the largest share of the post-transfer improvement in Mexico’s Gini coefficient after including programs targeting the poor (Esquivel et al, 2010).
E. Options to Strengthen Revenue Mobilization
20. To strengthen Mexico’s revenue mobilization, efforts could focus on broadening the tax base through the elimination of tax expenditures, combined with improvements in tax administration to enhance compliance. Mexico’s statutory rates appear broadly in line with key comparator countries, which suggest that priority should be given to efforts to widen the tax bases and strengthen administration. Table 3 illustrates a number of areas
24 In absolute terms the richest deciles spend more on consumption, and thus receive a larger share of the associated tax expenditure. But as a proportion of the income of each income group, the incidence of the implicit VAT subsidy is larger for the most vulnerable groups (OECD, 2011). Options for changing the system to broaden the tax base should consider associated measures to compensate these groups (as their purchasing power would be significantly reduced under the elimination of special VAT tax regimes).
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
35
I II III IV V VI VII VIII IX X
Households' income decile
Zero-rated VAT VAT exempted Fuels Residential electricity
Figure 13. Selected Implicit Subsidies by Income Decile(in percent)
Sources: based on SHCP (2011a)
41
where tax collection can be strengthened (to achieve levels closer to comparator countries) and presents indicative estimates of potential revenue gains.25 Key measures include:
VAT: reduced rates and exemptions should be revisited with the aim of eliminating them over time—in particular, the lower border rate is a concession not found elsewhere The lower bound of potential gains (from cutting tax expenditure) is estimated at 1.3 percent of GDP;26 more ambitious efforts could bring about 2.3 percent of GDP in additional revenues (to achieve a VAT revenue ratio closer to the OECD average).27
Environmental taxes: subsidies to gasoline and diesel should be eliminated over time, a measure that would increase revenue collection by about 0.6 percent of GDP. In addition, if Mexico were to tax fuels at the low end of the EU27 level,28 Mexico’s environmental taxes would yield an additional 1 percent of GDP, which Mexico has done in the past.29
Subnational taxes: As regards property tax, updating land registers and property values while improving administration and tax enforcement at the local level appears necessary to mobilize subnational revenue. This would help reach immovable property tax revenues closer to the LAC5 average (0.4 percent of GDP) while setting the ground to achieve OECD average collection levels over time (1.1 percent of GDP). Other taxes and special regimes should also be strengthened to further increase subnational revenue—e.g., small taxpayers regime (REPECOS), the local gasoline tax (administered by states), and the local vehicle-ownership tax and payroll tax among others.30
25 The lower bound indicates first-round gains obtained through a number of identified measures (elimination of various tax expenditures, introduction of a positive excise on gasoline and improving administration of property taxes); the upper bound represents an indicative benchmark defined to reflect international standards (which in principle could be achieved through the introduction of broader tax measures).
26 This is roughly given by 0.13 percent of GDP due to the reduced rate in border zones; 0.28 percent of GDP due to VAT exemptions; and about 0.87 percent of GDP due to goods and services at zero rate.
27 Keen et al (2009) undertake similar estimations to determine a tentative upper-bound for VAT revenue in Mexico. They also discuss in detail the steps involved to increase the VAT efficiency ratio in Mexico to levels observed in peer countries. Reaching those levels of VAT efficiency over the longer term would involve the elimination of special treatments (for instance of suppliers to the Maquiladora sector) and the correction of various imperfections in the system (including in administration and compliance). 28 For instance, Spain levied about 1 percent of GDP through fuel taxes in 2010.
29 Over the period 1980–2005, Mexico’s fuel excises were on average about 1 percent of GDP.
30 The focus on immovable property taxes here is due to the fact that in most OECD countries they tend to explain a large share of subnational tax collection. In addition, they have efficiency-enhancing characteristics relative to other taxes owing to the immobility of the tax base, which make them appealing from a tax policy
(continued)
42
Elimination of other key tax expenditures:31 tax expenditures associated with the purchase of new vehicles and other personal deductions should be revisited; similarly, special regimes such as that for the Maquiladora sector could be phased out. These measures could tentatively yield 0.5 percent of GDP in additional revenue. If employment subsidies and special deductions for wages and salaries were reduced (as in Table 3 below), income tax revenues could increase by 1.1 percent of GDP.32
21. To alleviate potentially adverse effects of revenue mobilization efforts on the poor, well-targeted social protection measures could be considered. The revenue gains associated with the increase in tax revenue would allow some resources to be allocated to social programs to alleviate the negative effects of the reforms on the poor. Moreover, in the context of needed reforms to boost Mexico’s growth potential, increasing social spending with high social return would help provide the basis for higher growth.
perspective. They are thus less distortionary in terms of the allocation of resources in the economy, since they do not tend to affect investment-savings decisions, thereby helping preserve long-run growth. They are also progressive, since property values and income levels are positively correlated with income.
31 Tax expenditures that are not strictly justifiable in terms of efficiency or effectiveness should be eliminated over time, as they tend to generate important costs to society and unnecessary complexities to the tax system.
32 This estimate, however, would have to be adjusted to net out that part of the revenue already collected through the IETU.
Measures to gain fiscal space
Value added tax: 1.3 - 2.6
elimination of reduced rates and exemptions 1/ 1.3 - 2.6
Environmental taxes: 1.7 - 1.7
elimination of fuel subsidies 2/ 0.6 - 0.6
fuel tax collection at low-end of EU27 3/ 1.1 - 1.1
Property taxes: 0.2 - 0.9
increase in immovable property taxes revenue 4/ 0.2 - 0.9
Elimination of other key tax expenditures: 0.3 - 1.1
personal income tax expenditures 5/ 0.1 - 0.5
corporate income tax expenditures 6/ 0.2 - 0.6
Total 3.5 6.2
1/ lower bound: SHCP 2010 estimates of VAT tax expenditure; upper bound: average 2008 revenue ratio of OECD. 2/ estimatation of 2010 fuel subsidy.3/ fuel taxes observed in Spain in 2010.
Source: Secretaría de Hacienda y Crédito Público (SHCP) and IMF staff calculations.
6/ lower bound: SHCP 2010 estimates of PIT tax expenditure on special regimes for Maquiladoras and vehicles up to Mex$ 175.000; upper bound: includes reduction of tax expenditure on employment subsidy and other CIT deductions.
Table 3. Quantification of options to increase non-oil tax revenue (in percent of GDP)
Revenue collection gains (lower and upper bounds)
4/ lower bound: to reach 2009 LAC5 average level; upper bound: to reach 2009 OECD average level.
5/ lower bound: SHCP 2010 estimates of PIT tax expenditure on various personal deductions; upper bound: includes reduction of various tax expenditures on wage income.
43
REFERENCES
Escolano, J., T. Komatsuzaki, B. Shang, and M. Soto, 2012, “Long-term projections of public spending and revenue for the pension and health system,” IMF Technical Assistance Report (Washington).
Esquivel, G., N. Lustig, and J. Scott, 2010, “Mexico: A Decade of Falling Inequality: Market
Forces or State Action?,” Unpublished Manuscript. http://www.noralustig.org/wp-content/uploads/2011/08/Esquivel-Gerardo-Nora-Lustig-John-
Scott.A-Decade-of-Falling-Inequality-in-Mexico-Market-Forces-or-State-Action.Chapter-7.pdf
Gómez Sabaini, J. C., and J. P. Jiménez, 2011, “Tax structure and tax evasion in Latin
America,” Serie Macroeconomía del Desarrollo, No, 118. Keen, M., G. Palomba, and Ch. McPherson, 2009, “Mexico: Selected Tax Reform Issues,”
IMF Technical Assistance Report (Washington). Lopez-Murphy, Pablo, 2011, “Long-Term Fiscal Challenges in Mexico,” Selected Issues
Paper, IMF Country Report No. 11/249, pp. 27–38. OECD, 2010a, Tax Policy Reform and Economic Growth, OECD Publishing. OECD, 2010b, Choosing a Broad Base–Low Rate Approach to Taxation, OECD Publishing. OECD, 2011, Economic Surveys: Mexico, OECD Publishing. SHCP (Secretaría de Hacienda y Crédito Público), 2011a, “Distribución del pago de
impuestos y recepción del gasto público por deciles de hogares y personas: Resultados para el año 2010,” SHCP Publishing.
SHCP (Secretaría de Hacienda y Crédito Público), 2011b, “El Impuesto Empresarial a Tasa
Única (IETU): Un diagnóstico a tres años de su implementación,” SHCP Publishing. Villela, L., A. Lemgruber and M. Jorratt, 2010, “Tax Expenditure Budgets,” IDB Working
Paper No. IDB-WP-179.
44
Appendix 1. Mexico: Breakdown of Tax Expenditures in 2008, 2009, and 2010
Mex$ bn percent of GDP Mex$ bn percent of GDP Mex$ bn percent of GDP
Total 629 5.2 335 2.8 375 2.9Income taxes 1/ 206 1.7 159 1.3 168 1.3
o/w special deductions of Maquiladoras (ISR empresas) 13 0.1 13 0.1 14 0.1o/w deductions for vehicles up to Mex$ 175,000 (ISR empresas ) 9 0.1 17 0.1 17 0.1o/w employment subsidy (ISR empresas) 35 0.3 32 0.3 34 0.3o/w various wage exemptions (ISR personas fisicas) 67 0.6 46 0.4 49 0.4o/w other deductions (ISR personas fisicas) 26 0.2 9 0.1 9 0.1
VAT 210 1.7 159 1.3 167 1.3Zero-rated 2/ 152 1.2 108 0.9 114 0.9Exempted 3/ 43 0.4 34 0.3 36 0.3Reduced rate at the border (11 percent) 15 0.1 16 0.1 16 0.1
Special taxes 199 1.6 7 0.1 34 0.3o/w negative IEPS 4/ 196 1.6 6 0.0 32 0.2
Fiscal stimulus 14 0.1 10 0.1 6 0.0
Memorandum items:Nominal GDP (Mex$ bn) 12176 12176 11930 11930 13084 13084IETU tax expenditure (percent of GDP) 5/ 0.8 0.8 0.7 0.7 0.6 0.6IETU tax collection (percent of GDP) 0.4 0.4 0.4 0.4 0.3 0.3
Sources: Secretaría de Hacienda y Crédito Público (SHCP): Presupuesto de Gastos Fiscales, various issues, and IMF staff calculations.
1/ excludes tax expenditures and tax collection of IETU.
2/ includes: food, medicines, books and newspapers, imported goods by Maquiladoras, hotels services used for conventions, and other goods and services.
3/ includes: medical, educational and transport services, cultural events and housing.
4/ may differ from actual figures due to differences between projections and outturns.
5/ as reported by Mexican authorities in Presupuesto de Gastos Fiscales, various issues.
2008 2009 2010
45
IV. MEXICO: FINANCIAL SECTOR DEVELOPMENT1
While Mexico’s bond and equity markets have developed rapidly during recent years, banking sector credit has remained relatively subdued compared to emerging market peers. Evidence suggests that bond and equity financing in Mexico can be explained by usual factors (economic development, saving rate, sovereign debt market depth, institutional/regulatory environment), but bank financing has remained below what those factors would suggest. The sustained increase of bank credit after the global crisis has helped to start reversing this puzzle, with potential implications for access to credit for small and medium size enterprises and growth.
A. Background
1. There is growing consensus in the literature that financial development matters for economic growth. Since the early work by Diamond (1969) on the relationship between financial development and growth, substantial work has been done, particularly during the last 20 years.2 Financial development can have an impact on economic growth through different channels, including by (i) facilitating diversification of risks; (ii) mobilizing savings; (iii) reducing the cost of external financing for firms; and (iv) disseminating information and monitoring investment decisions. The empirical literature also suggests that financial depth is important for growth.3
2. The structure of the financial system also matters, as it may affect access to credit for some companies, in particular small and medium size enterprises (SMEs). Although the literature has not found a direct link between the relative development of bank-based and market-based financing and economic growth, it is generally acknowledged that bond and equity financing is less effective in serving borrowers in situations characterized by “opaqueness” (i.e., the difficulty to ascertain whether firms have the capacity or the willingness to pay). Banks, through a number of transactional technologies, are better suited to serve companies for which the lender has to rely predominantly on soft information.4
1 Prepared by Francisco Vázquez Ahued and Esteban Vesperoni.
2 See, for example, Beck et.al. (2001), Goldsmith (1969), Levine (1997 and 2005), Rajan and Zingales (1998), Stultz (2001), and references therein.
3 For a summary on theoretical and empirical underpinnings in the relationship between financial development and growth, see Levine (2005).
4 These include relationship lending, credit scoring, standardized risk rating tools, asset-based lending, factoring, leasing, etc. See, for example, Berger and Udell (2006), and de la Torre, Martinez Pería and Schmukler (2010). Other papers on these issues are Armendariz et. al. (2005), Beck, Demigurc-Kunt and Martinez Pería (2008 and 2010); and Beck and Demigurc-Kunt (2007).
46
3. The development of domestic bond and equity markets has helped Mexico recover financing levels reached during the early 1990s. Total credit to the private sector (including bank credit as well as bond and equity financing), which peaked at more than 50 percent of GDP in 1994 (Figure 1), fell to less than 40 percent in the early 2000s and has only recently recovered past high levels (data sources for Figures 1–4 included in Appendix Table 1). This recovery has been mainly driven by the gradual expansion of bond and equity markets between 1994 and 2010 (Figures 2a and 2b), which has likely been driven in part by the reform of the pension system.5 Domestic bond markets, in particular, have grown significantly and reached market capitalization levels of around 20 percent of GDP, larger than the average in other EMs. Equity markets have also developed in line with other EMs, although equity financing is still lower than in comparator countries (this financing includes the cumulative value of annual Initial Public Offerings, as percent of GDP in the year of issuance).6 Despite the development of market financing, total financing in Mexico still appears low compared to other EMs, including other Latin American economies.
5 Assets under management by pension funds represented more than 10 percent of GDP by 2010.
6 These comparisons are based on a group of EMs in Latin America, Asia and Europe that have depicted dynamic financial systems during the last 20 years. The sample includes Brazil, Bulgaria, Chile, Colombia, Czech Republic, Estonia, Hungary, Indonesia, Korea, Latvia, Lithuania, Malaysia, Mexico, Peru, Philippines, Poland, Romania, Slovak Republic, Thailand, Turkey, and South Africa. In Figure 1, rest of the sample refers to non Latin American countries listed above.
0
2
4
6
8
10
12
14
16
18
20
0
2
4
6
8
10
12
14
16
18
20
MexicoLA5 excluding MexicoRest of sample
Figure 2a. Domestic Private Bonds (market capitalization, share of GDP)
5
10
15
20
25
30
35
5
10
15
20
25
30
35MexicoLA5 excluding MexicoRest of sample
Figure 2b. Equity(equity financing, share of GDP)
20
40
60
80
100
120
140
20
40
60
80
100
120
140MexicoLA5 excluding MexicoRest of sample
Figure 1. Total Private Sector Financing (share of GDP)
47
4. Bank credit has remained relatively subdued (Figure 3). After reaching almost 40 percent of GDP in 1994, bank credit to the private sector (defined as claims to the private sector by other depositary corporations, as reported in IFS) hit a trough at about 13 percent of GDP in 2004. By 2010, bank financing had rebounded to about 20 percent of GDP, still well below previous peaks.7 Bank credit to GDP also represents less than half the level of other EMs in the sample. More recently, bank credit has shown a sustained recovery, growing at about 10 percent in real terms during the last couple of years.
5. As a result of these trends, the financial sector in Mexico has become less bank-based, compared both to the past and to other EMs (Figure 4). The share of bank credit in total financing to the private sector, which reached almost 75 percent in 1994, was about 35 percent in 2010.8 This contrasts with the experience of other EMs, in which the share of bank credit has remained at around 50–60 percent during the last 20 years.
7 This performance in banking system credit is consistent with the low level of intermediated funds. Bank deposits peaked during the 1990s at more than 30 percent of GDP, a level that was broadly similar to other EMs; and in particular to Latin American peers. However, the trend in deposits reversed in the mid-1990s, and bottomed out at about 20 percent of GDP in 2004. Deposits recovered to about 27 percent of GDP by 2010, but it is still well below the average level in other EMs (57 percent of GDP). Loan-to-deposit ratios in Mexico—at 1.2—are also lower than the average in other EMs at 1.65. Relatively low leverage may have also contained credit growth, although it has likely shielded the banking system from vulnerabilities experienced in other EMs.
8 The share of bank credit bottomed in 2004 at about 30 percent.
10
20
30
40
50
60
70
80
10
20
30
40
50
60
70
80Mexico
LA5 excluding Mexico
Rest of sample
Figure 3. Bank Credit to Private Sector(share of GDP)
30
40
50
60
70
80
30
40
50
60
70
80MexicoLA5 excluding MexicoRest of sample
Figure 4. Banking Credit (share of total private financing)
48
B. Financial Development: Mexico in the Context of Emerging Markets
6. To study financial sector development in Mexico a cross-country econometric analysis was performed on the drivers of bank- and market-based financing between 1990s and 2010 (Appendix 1). The main results show that:
The evolution of market-based financing in Mexico is explained by usual fundamentals, in line with other EMs. Regression analysis suggests that the dynamics of market financing is driven by the size of the economy, saving rates, the development of sovereign debt markets, institutional factors, foreign financing spreads, and the cost of equity financing. Moreover, market-based financing is usually higher after financial crises, with Mexico’s well captured by the regression analysis without a significant fixed effect.
The evolution of bank credit in Mexico, however, remains only partly explained by usual fundamentals. The analysis suggests that the factors explaining market-based financing help also understand bank credit in EMs. However, while these variables help explain market-based financing in Mexico, the specification for bank financing cannot reject a significant negative fixed effect, suggesting that that macroeconomic and financial factors would predict higher bank credit in Mexico.9
A sequential analysis of crises suggests that the protracted stagnation of bank credit in Mexico is not likely associated with the direct impact from the 1994 crisis. Figure 5 (appendix) shows that a crisis dummy for Mexico becomes significant only in 2005, a decade after the 1994 financial crisis, suggesting that other factors not associated with the immediate financial distress may be at play in the evolution of bank credit, including a lower level of savings through bank deposits and the implementation of tighter prudential regulations that usually follow financial crises.10
9 Although beyond the scope of this note, other factors at play could be related to competition in the banking sector.
10 For some countries, notably Mexico, pension system reforms after financial crises may have also played a role.
49
REFERENCES
Armendariz de Aghion, B. and J. Morduch, 2005, “The Economics of Microfinance” Cambridge, MA: MIT Press.
Beck Thorsten, A. Demigurc-Kunt, and M. Martinez Peria, 2008, “Banking Services for Everyone? Barriers to Bank Access and Use around the World” The World Bank Economic Review, Volume 22 (3), 397-430.
___________, 2011, “Bank Financing for SMEs: Evidence Across Countries and Bank Ownership Types” Journal of Finance Services Research, 39, 35–54.
Berger, Allen, and Gregory Udell, 2006, “A More complete Conceptual Framework for SME Finance”, Journal of Banking and Finance, 30, 2945–2966.
Cull, Robert and Maria S. Martinez Peria, 2010, “Foreign Bank Participation in Developing Countries” World Bank Policy Research Working Paper 5398, World Bank.
De la Torre, A., A. Ize, and S. Schmukler, 2012, “Financial Development in Latin America and the Caribbean: The road Ahead” The World Bank.
De la Torre, A., M. Martinez Peria, and S. Schmukler, 2010, “Bank Involvement with SMEs: Beyond Relationship Lending” Journal of Banking and Finance, 34, 2280–2293.
Goldsmith, R., 1969, “Financial Structure and Development” New Heaven, CT: Yale University Press.
Levine, Ross, 1997, “Financial Development and Economic Growth: Views and Agenda” Journal of Economic Literature, Volume XXXV, 688–726.
___________, 2005, “Finance and Growth: Theory and Evidence” Handbook of Economic Growth, Volume 1A.
Levine, Ross, A. Berger, A. Demirguc-Kunt, and J.G. Haubrich, 2004, “Bank Concentration and Competition: An Evolution in the Making” Journal of Money, Credit, and Banking, 36(3), 433–452.
Rajan, Raghuram and Luigi Zingales, 1998, “Financial Dependence and Growth” American Economic Review 88, 559–586.
Stultz, R., 2001, “Does Financial Structure Matters for Economic Growth? A Corporate Finance Perspective” In: Demirgüç-Kunt, A., Levine, R. (Eds.), Financial Structure and Economic Growth: A Cross-Country Comparison of Banks, Markets, and Development. MIT Press, Cambridge, MA, pp. 143–188.
50
APPENDIX 1. CROSS-COUNTRY ANALYSIS OF FACTORS BEHIND FINANCIAL SYSTEM
DEVELOPMENT
This appendix analyzes the factors behind the evolution of bank- and market-based financing in EMs. Panel regressions comprising the EMs in the sample are used for the period 1990–2010.11 The dependent variables are: (i) the ratio of financing instruments issued in capital markets (i.e., bonds and equity) to GDP; (ii) the ratio of bank financing to the private sector to GDP; and (iii) the ratio of financing instruments issued in capital markets to total financing to the private sector.12 The method of estimation is GMM, including country dummies.13
Three specifications are presented. An initial specification regressing the dependent variables on development, macroeconomic and financial factors is followed by an analysis of the impact of financial crises. Finally, institutional factors are added to the specification.
Basic Specification
Bank credit and market-based instruments are affected inter alia by the level of development, savings, the depth of sovereign debt markets and funding costs (Table 2). The results in the basic specification suggest that: (i) both bank-based and market-based financing increase with the size of the economy; (ii) increasing levels of savings (i.e., total savings for market-based instruments and bank deposits for bank credit) are associated with higher financing; (iii) the development of sovereign debt markets helps boosting market-based instruments, but it has a negative impact on bank financing; (iv) countries with more open capital accounts, by facilitating access of domestic corporations to external credit markets, show lower levels of domestic market-based financing; (v) higher international interest rates and more expensive domestic equity markets are associated with lower market-based financing in EMs; and (vi) higher international interest rates are associated with more bank financing, which is likely funded to a large extent by domestic deposits.
11 Some of the countries in the sample have not been included in the different regressions depending on the availability of information, in particular for bond and equity market instruments. The list of variables is described in Table 1.
12 The variable on equity financing has been adjusted for valuation changes, i.e. the variable captures the net financing associated to new issuances at a given year, valued in terms of the GDP of that year.
13 The following variables have been instrumented by using their lagged values: GDP per capita, domestic savings, public bond market capitalization, and capital account openness.
51
The role of Crises
After financial crises, countries seem to experience a reduction in bank credit concomitantly with an increase in market-based financing. In order to test the effect of financial crises on the depth and structure of EMs credit markets, a step-dummy variable is introduced to the basic specification.14 The results suggest that after financial crises, EMs experience a reduction in bank credit and an increase in market-based financing. Financial crises, however, usually bring about significant changes in the functioning of banking systems and in regulatory frameworks, which cannot be easily disentangled.
A sequential analysis suggests that the dynamics of bank credit after financial crises is likely associated with a number of factors.15 The basic specification incorporates a crisis-dummy sequentially.16 The results suggest that the impact of crises is not necessarily related to the financial distress itself, but to changes that take place following the crises, (Figure 5 illustrates the results of the analysis for some EMs, including Mexico). Many factors may explain changes in banking sector credit after financial crises, including inter alia a lower level of savings in the form of bank deposits and the implementation of tighter prudential regulations.
14 The variable takes the value zero as long as a country did not experience a financial crisis, and one afterwards.
15 The analysis was also performed for market-based financing instruments, with similar results.
16 Starting with a single 1-value dummy variable during the crisis year, we run multiple regressions incorporating an extra 1-value observation in subsequent years in each regression.
‐15
-10
-5
0
5
10
-10 -5 0 5 10 15 20
Cri
sis
Du
mm
y
Country Fixed Effects
Appendix 1. Bank Credit to the Private Sector (Share of GDP)
Mexico
Brazil
Turkey2005
2006
2007
20082009
2010
Red dots represent crisis dummy significant at least at the 10% level of confidence
2000
2001
2010
52
Institutional Factors
An institutional variable is added to the specification. The World Bank’s World Governance Indicator (WGI) for rule of law is used, which captures perceptions of the extent to which agents have confidence in and abide the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence. The WB produces a score based on an unobserved components model for each economy, and then ranks them in a scale of 0 to 100.17
Better institutions have a positive effect on bank lending in EMs (Table 4). The new specification shows that the WGI variable is positive and significant in the regression of bank credit in EMs, which is likely related to an increase in the quality of contract enforcement and property rights. The crisis variable, however, becomes non-significant suggesting that the crisis step-dummy variable may be capturing an effect not directly related to financial distress during crises.
17 The World Bank produces this indicator bi-annually between 1994 and 1998. We interpolate the values of the indicator for 1995 and 1997.
53
Appendix Table 1. Summary of Variables
Variables Source(s) Units
GDP per capita WEO, Sept. 2011 USD
National savings as percent of GDP WEO, Sept. 2011 Percent
Government domestic bonds market capitalization as percent of GDP
BIS, Quarterly Statistics Table 16B Percent WEO, Sept. 2011
Private domestic bonds market capitalization as percent of GDP
BIS, Quarterly Statistics Table 16A Percent WEO, Sept. 2011
Capital Account openness Chinn Ito Financial Openness Index (normalized)
Between 0 and 1
Deposits as percent of GDP IFS Percent
Bank credit IFS Percent
Equity market financing as percent of GDP S&P Global Stock Markets Factbook, 2011 Percent Dealogic
Equity markets index S&P Global Stock Markets Factbook, 2011 Index, 2000=100
U.S. 2-Years Treasury Bond Haver Analytics Percent World Bank Governance Indicator World Bank Percentile
Variable Variables used to construct, and
source(s), if not defined above Definition
Capital markets as percent of GDP pvtbond sm_adjusted Nominal GDP (in USD): WEO, 2011
_100
Capital markets as percent of total private financing
Pvtbond sm_adjusted creditpvt
__
100
54
Appendix Table 2. Bank and Market-Based Financing: Basic Specification
(1) (2) (3) Market-based
financing as share of GDP
Banking financing as share of GDP
Market-based as share of total
financing GDP per capita in USD 0.0022*** 0.0024*** 0.0000 Domestic Savings 0.8958*** 0.8313** Public Bond Market Capitalization
0.2735*** -0.5146*** 0.4197***
Capital Account Openness
-10.9993** -4.4179 0.4574
T-Bond, 2 YRS -1.3637*** 0.8655** Stock Market Index -0.0079*** Banking Institutions: Deposits
0.5173*** -0.0620
Observations 255 278 251 Standard errors in parentheses * p<.10, ** p<.05, *** p<.01
55
Appendix Table 3. Bank and Market-Based Financing: Crises (1) (2) (3) Market-based
financing as share of GDP
Banking financing as share of GDP
Market-based as share of total
financing GDP per capita in USD 0.0017*** 0.0025*** -0.0002 Domestic Savings 0.7398** 0.6774** Public Bond Market Capitalization
0.1993** -0.4550*** 0.3170***
Capital Account Openness
-7.3950* -7.2752* 4.0016*
T-Bond, 2 YRS -0.9868*** 0.6888** Stock Market Index -0.0060*** Crises 7.0325*** -6.6633** 8.0622*** Banking Institutions: Deposits
0.6596*** -0.2296***
Observations 255 278 251 Standard errors in parentheses * p<.10, ** p<.05, *** p<.01
56
Appendix Table 4. Bank and Market-Based Financing: Institutional Factors (1) (2) (3) Market-based
financing as share of GDP
Banking financing as share of GDP
Market-based as share of total
financing GDP per capita in USD 0.0011*** 0.0023*** -0.0001 Domestic Savings 0.5115* 0.1529 Public Bond Market Capitalization
0.3389*** -0.6030*** 0.3572***
Capital Account Openness
-0.3988 -9.6663 5.1536*
T-Bond, 2 YRS -0.9338*** 0.3940 Stock Market Index -0.0054*** World Bank Governance Indicator
0.1453 0.4950*** -0.1143
Crises 2.9188* -4.0389 4.4026*** Banking Institutions: Deposits
0.5259*** -0.2330***
Observations 222 239 220 Standard errors in parentheses * p<.10, ** p<.05, *** p<.01
57
V. MEXICO: MIGRATION AND LABOR MARKETS1
Urban unemployment in Mexico remains higher than in past recoveries. At the same time, Mexican migration to the United States has decelerated significantly since the mid-2000s, linked to the slump in construction in the U.S. This suggests that the dynamics of migration has played a role in the evolution of urban unemployment in Mexico through an increase in participation in the labor force. In turn, the increase in labor participation has contributed to contain labor costs and to help regain competitiveness. Over the medium term, the effects of migration on urban unemployment and output would depend on the recovery in the U.S. construction sector and on how permanent the change in migration dynamics is. The labor reform being considered by congress would facilitate adjustments in labor markets in coming years.
A. Background
1. The level of urban unemployment in Mexico is falling after the crisis, but remains elevated compared with the early 2000s, when it reached a multi-decade low. After peaking at more than 7½ percent in 1995, urban unemployment fell steadily following the recovery from the 1994 crisis and the incorporation of Mexico into NAFTA. As a consequence, urban unemployment hit its minimum since the mid-1980s in the last quarter of 2000. In the last decade, the evolution of urban unemployment could be characterized by three different plateaus (Figure 1). The early 2000s were characterized by very low urban unemployment rates, which trended up after 2003 into a ‘mid-decade plateau’. A third plateau with urban unemployment around 6 percent followed the fall in economic activity after the global crisis.
2. Mexican migration to the United States decelerated in the 2000s compared to previous decades. Migration to the U.S. was particularly dynamic during the 1970s and 1980s.2 However, recent data from both INEGI and the American Community Survey (ACS) shows that migration flows have decelerated sharply during the second half of the 2000s,
1 Prepared by Enrique Flores and Esteban Vesperoni.
2 See CONAPO at http://www.conapo.gob.mx/
2
3
4
5
6
7
8
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Figure 1. Urban Unemployment(In percent)
Sources: INEGI, serie unificada
58
associated with the weak U.S. construction sector and the increase in unemployment in U.S. states with a strong presence of Mexican population.
B. Recent Developments in Labor Markets
3. Following the economic recovery after the 1994 crisis, Mexico recorded historically low unemployment rates in the early 2000s. After a prolonged and sustained fall since the mid 1990s, urban unemployment rate reached 2¾ percent in December 2000, the lowest level on record since the mid 1980s. This historically low urban unemployment rates took place in the context of fairly stable participation rates and moderate employment creation during this period (Table 1).3
4. An increase in labor participation before the global crisis pushed up urban unemployment despite buoyant labor demand. The four years before the global crisis showed a significant acceleration in employment creation (from 1½ percent to 2¼ percent). Despite buoyant labor demand, the urban unemployment rate in this period increased to 4¾ percent, from slightly above 4 percent in 2000–2004, due to the increase in participation rates since 2003. This latter peaked at over 61 percent at end-2007 (Figure 2).
3 Labor participation, which had stood at around 50 percent in the late 1980s/early 1990s, increased significantly during the first five years before the Tequila crisis, not only pushed by favorable cyclical conditions, but also in part by an increase in labor participation by women. After reaching values around 55 percent in 1993/4, it remained stable during the following 10 years to the early 2000s. For an analysis of labor participation in Mexico, see Duval-Hernández and Orraca Romano (2009).
Period Labor Employment
Participation 2/ 3/
2000-2004 1.86 1.51
2005-2008 2.16 2.21
2008-2009 1.46 0.24
2010-2012 1.91 1.56
2011-2012 2.50 2.73
Source: INEGI (ENOE).
1/ Year-on-year growth rates, at a quarterly frequency.
2/ Number of people in the economic active population.
3/ Number of people employed.
Average Quarterly Growth 1/
Table 1. Mexico: Labor Market Indicators
59
5. The global crisis halted job creation and triggered a sharp increase in urban unemployment. While labor participation stabilized at around 60 percent of the working age population during the global crisis, the creation of employment plummeted, falling to about ¼ percent from about 2¼ percent in the period 2005–2008. In turn, the urban unemployment rate spiked to more than 7 percent in 2009.
6. Urban unemployment has fallen with the recovery in economic activity in 2010-2012, but it has remained high compared to historical averages. Urban unemployment remains at 5½-6 percent, compared with 4¾ percent before the global crisis. Despite that employment creation has been strongest since the early 2000 (see Table 1), labor participation has continued to growth more rapidly than in the past, especially in 2011 and 2012.
7. The increase in labor participation has likely contributed to a reduction in unit labor costs (Figure 3). Despite a healthy expansion in employment between 2005 and 2008, unit labor costs in the manufacturing sector remained broadly stable. Following the global crisis, manufacturing unit labor costs have fallen by about 5 percent between end-2009 and end-2011.4
C. Migration: Recent Developments
8. The number of Mexican migrants to the U.S. grew rapidly during the 1970–2000 period. In 1970, there were close to one million Mexican born residents in the U.S. That number increased dramatically to almost 9 million in 2000, after the strong Mexican migration flows in previous decades. These migrants were mainly located in California (about 45 percent), Texas (over 20 percent), Arizona, and Illinois.
4 The dynamics of wages—and unemployment—may also be affected by participation in formal and informal labor markets. See Alcaraz (2009) and Alcaraz, Chiquiar and Ramos Francia (2008).
85
90
95
100
105
110
115
120
Figure 3. Unit Labor Costs(Index: 2008=100)
Sources: INEGI and staff calculations
2.00
4.00
6.00
8.00
55
60
65
2000 2001 2002 2003 2004 2005 2006 2007 2008 2010 2011 2012
Participation (LHS)Unemployment (RHS)
Figure 2. Unemployment and Participation(In percent)
Sources: INEGI, series unificadas
60
9. Migration flows to the U.S. fell significantly in the 2000s, particularly after 2005. Migration flows decelerated sharply in the second half of the 2000s, with some indicators showing reversed migration flows from the U.S. to Mexico during the global crisis:5
The American Community Survey (ACS) is an ongoing statistical survey that presents data on Mexican-born residents in the U.S. since 2000. The ACS shows that these residents increased by 35 percent between 2000 and 2010; it also shows that the net number of Mexican migrants to the U.S. was negative in 2008 (Figure 4a).6
A survey by INEGI compiling quarterly information on net migration flows suggests that these flows have declined sharply since 2006 (from about 144 migrants per ten thousand people in Q2-2006 to about 30 in Q4-2011) and has stabilized at low levels in the last few years (Figure 4b).7
10. The data suggests that migration began decelerating with the downturn in the construction sector in the U.S. Annual data from the ACS suggests that the turning point in U.S. construction activity and migration from Mexico to the U.S. took place in early 2006 (Figure 5), after which unemployment rates in some key states in the U.S. began to increase.8
5 Data on migration flows between Mexico and the U.S. can be obtained from different sources, but all of them offer a similar picture for the last decade.
6 For a comprehensive analysis of this data, see Cervantes González (2011).
7 In contrast with the ACS, data from INEGI does no show a recovery in migration flows after the global crisis (Figure 5).
8 Quarterly migration data from INEGI and regional unemployment rates in the U.S. also support this point.
-400
-200
0
200
400
600
800
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Figure 4a. Mexico: Migration to the U.S.(Thousand per year)
Sources: American Community Survey
20
40
60
80
100
120
140
160
Figure 4b. Migration to the U.S.(Migrants per ten thousand people)
Sources: INEGI (ENOE)
-2.00
-1.50
-1.00
-0.50
0.00
0.50
1.00
1.50
-4
-2
0
2
4
6
8
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Migration (% change, lhs)
US Construction (% change)
Figure 5. Migration and U.S. Construction(In percent)
Sources: ACS, U.S. Department of Commerce
61
D. Migration and Labor Markets
11. Evidence suggests that migration has had an impact on urban unemployment during the last decade. The impact of migration flows on urban unemployment has in part taken place through the increase in labor participation in Mexico in the second half of the 2000s (Figure 6). The sustained increase in participation has also played an important role in the sluggish fall in urban unemployment rates that took place after the global crisis, despite a healthy recover in economic activity and employment creation in Mexico.
12. The longer-term impact of migration on urban unemployment (and potential output) will depend on how persistent the recent trends in migration are. In particular, the effect of migration on urban unemployment would depend on: (i) the pace and extent of the recovery in the U.S. construction sector, which is still 35 percent below its peak; (ii) the dynamics of migration flows as the U.S. construction activity recovers; and (iii) the impact of higher labor participation on real wages in Mexico. A permanent shift in labor participation in Mexico can potentially have a sustained impact on unit labor costs and increase potential output in Mexico.
13. Prospects for a persistent increase in Mexico’s labor participation highlight the importance of labor market flexibility. In this context, the labor reform being considered by congress—which allows new types of temporary employment contracts, streamlines the settlement of labor disputes and limit compensation for unjustified dismissals, introduces more flexible seniority rules favoring productivity and labor skills, and changes the regulatory framework for outsourcing—would be important to facilitate the incorporation of new entrants in the labor markets.9
9 For the effects of labor market rigidities on productivity, see Chiquiar and Ramos Francia (2009).
58
59
60
61
62
20
40
60
80
100
120
140
160
2006 2007 2008 2009 2010 2011
Migration (per thousand people, lhs)Labor Participation (-1)
Figure 6. Labor Participation and Migration
Sources: INEGI (ENOE)
62
REFERENCES
Alcaraz, Carlo, 2009, “Informal and Formal Labor Flexibility in Mexico,” Desarrollo y Sociedad, Primer Semestre de 2009, pp 115–143.
Alcaraz, C. and D. Chiquiar, 2008, “Diferenciales salariales intersectoriales y el cambio en la composición del empleo urbano de la economía Mexicana en 2001–2004,” Banco de Mexico Working Paper 2008–2006.
Cervantes González, Jesús, 2001, “Remesas familiares y la migración de mexicanos a Estados Unidos,” Centro de Estudios Monetarios Latinoamericanos, Documento de Trabajo.
Chiquiar, D. and M Ramos Francia, 2009, “Competitiveness and Growth of the Mexican Economy,” Banco de Mexico Working Paper 2009–2011.
Duval-Hernández, R. and P. Orraca Romano, 2009, “A Cohort Analysis of labor Participation in Mexico, 1987–2009,” IZA Discussion Series Paper 4371.
![what works? [ 2 ] v21](https://img.pdfslide.us/doc/110x75/6284c299450b3950ec18acfe/what-works-2-v21.jpg)
